Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.

org/
 Seismic Data Processing
with Seismic Un*x
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

A 2D Seismic Data Processing Primer

David Forel, Thomas Benz,

and Wayne D. Pennington
Michigan Technological University

Course Notes Series No. 12

Lawrence Gochioco, Series Editor

Society of Exploration Geophysicists

Tulsa, Oklahoma, USA
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

SEG Policy on Software-Related Books

SEG encourages the submission of manuscripts on software-related topics for publication

in print and/or electronic form if the work’s primary aim is to promote the understanding
of geophysics and if the work involves geophysical software that is either free or in wide
distribution. SEG is not interested in publishing manuals describing only the technical
aspects of operating geophysical software. However, SEG does seek to publish self-
guided tutorials and other treatises focused on explicating geophysical principles that may
heavily involve specific free or widely used geophysical software. SEG does not promote
the discussed software to the exclusion of any other product.

About the SEG Course Notes Series

SEG Course Notes Series publications are published without the normal SEG peer
reviews. This volume in the series is reproduced here as provided by the authors.

ISBN 978-0-931830-48-8 (Series)

ISBN 978-1-56080-134-4 (Volume)

Society of Exploration Geophysicists

P.O. Box 702740, Tulsa, OK 74170-2740, USA

© 2005 by the Society of Exploration Geophysicists

All rights reserved (except as noted in text)

Published in 2005
Reprinted in 2007, 2008

Printed in the United States of America

 Short Table of Contents

Short Table of Contents ........................................................................................... iii

Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Table of Contents .................................................................................................... v

Preface...................................................................................................................... xi
Table of Computer Notes ......................................................................................... xii
Table of Scripts ........................................................................................................ xiii
1. Introduction .......................................................................................................... 1-1
2. Unix Commands and Concepts............................................................................ 2-1
3. Trace Headers and Windowing Data ................................................................... 3-1
4. Three Simple Models ........................................................................................... 4-1
5. Three Simple Models: Acquire 2-D Lines ........................................................... 5-1
6. Model 4: Build, Acquire a Line, Display Gathers, QC ........................................ 6-1
7. Model 4: Sort, Velocity Analysis ......................................................................... 7-1
8. Model 4: T-V Picks QC, NMO, Stack ................................................................. 8-1
9. Model 4: Migration .............................................................................................. 9-1
10. Nankai Data: Examine, Resample, Sort, Gain ................................................... 10-1
11. Nankai: Velocity Analysis, NMO, Stack, Migration .......................................... 11-1
12. Taiwan Data: Examine, Zero Traces, Re-examine ............................................. 12-1
13. Taiwan: Gain-Filter, Filter-Gain ........................................................................ 13-1
14. f-k Filter and Deconvolution.............................................................................. 14-1
15. Taiwan: Sort, Velocity Analysis, NMO, Stack ................................................... 15-1
16. Taiwan: Migration .............................................................................................. 16-1
References ................................................................................................................ R-1
Appendix A. Seismic Un*x Legal Statement .......................................................... A-1
Appendix B. Seismic Un*x at Michigan Tech ......................................................... B-1
Appendix C. Utility Programs ................................................................................. C-1
Appendix D. Makefiles: Alternative to Shell Scripts, by Chris Liner ..................... D-1
Appendix E. On the CDs ......................................................................................... E-1

iii
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

This page has been intentionally left blank

 Table of Contents

Short Table of Contents ........................................................................................... iii

Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Table of Contents ..................................................................................................... v

Preface ...................................................................................................................... xi
Table of Computer Notes ......................................................................................... xii
Table of Scripts ........................................................................................................ xiii
1. Introduction ...................................................................................................... 1-1
1.1 The Goal of this Primer .............................................................................. 1-1
1.2 The Outline of this Primer .......................................................................... 1-1
1.3 A Bit about Seismic Un*x .......................................................................... 1-2
1.4 Computer Note 1: Enter ............................................................................. 1-2
1.5 Computer Note 2: Your Shell ..................................................................... 1-3
1.6 Computer Note 3: Changing Directories .................................................... 1-3
1.7 Computer Note 4: Cursor Prompt .............................................................. 1-3
1.8 Your First Seismic Un*x Instruction .......................................................... 1-4
1.9 Computer Note 5: Background (&) and Foreground Processes ................. 1-5
1.9.1 Background Processes .......................................................................... 1-5
1.9.2 Foreground Processes ........................................................................... 1-6
1.10 Zoom the xwigb Window ........................................................................... 1-6
1.11 The Demos Directory ................................................................................. 1-7
1.12 Self Documentation in Seismic Un*x ........................................................ 1-7
1.13 Your Second Seismic Un*x Instruction ..................................................... 1-8
1.13.1 Computer Note 6: No Spaces around the Equal Sign .......................... 1-9
1.14 Other Help Facilities in Seismic Un*x ....................................................... 1-9
2. Unix Commands and Concepts ........................................................................
2.1 Elementary Unix Commands ..................................................................... 2-1
2.2 File Name Conventions .............................................................................. 2-2
2.3 Advanced Unix Concepts ........................................................................... 2-2
2.3.1 Invoke the Shell: #! /bin/sh .......................................................... 2-2
2.3.2 Comment Line: # .................................................................................. 2-2
2.3.3 Pipe: | .................................................................................................... 2-2
2.3.4 Shell Redirection: < and > and >> ....................................................... 2-3
2.3.5 Permissions: rwxrwxrwx ...................................................................... 2-3
2.3.6 Shell Internal Variables ........................................................................ 2-3
2.3.7 Passing Values to the Shell .................................................................. 2-4
2.3.8 Continuation Mark: \ ............................................................................ 2-4
2.3.9 Integer Evaluation ................................................................................ 2-4

v
TOC-1
 Table of Contents

2.3.10 Floating Point Evaluation ..................................................................... 2-4

2.3.11 Debug Option ....................................................................................... 2-5
2.3.12 Case Option .......................................................................................... 2-5
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

2.3.13 Shell Loops and If Blocks .................................................................... 2-5

2.3.14 Exit the Shell ........................................................................................ 2-5
2.4 Shell Script Example: myplot.sh ................................................................ 2-5
3. Trace Headers and Windowing Data ................................................................
3.1 What are Trace Headers? ........................................................................... 3-1
3.2 First Look at Trace Headers (keys): surange and sugethw ....................... 3-1
3.3 Real Seismic Shot Gathers ......................................................................... 3-2
3.3.1 Get the Data .......................................................................................... 3-3
3.3.2 Test the Data ......................................................................................... 3-3
3.3.2.1 Big-Endian (High Byte) Machine: Solaris ..................................... 3-3
3.3.2.2 Little-Endian (Low Byte) Machine: PC ......................................... 3-3
3.4 Program suxwigb and Option perc ............................................................ 3-5
3.5 Second Look at Trace Headers (keys): sukeyword ................................... 3-5
3.6 Third Look at Trace Headers (keys): suxedit ............................................ 3-6
3.7 Trace Header Manipulation: Constant and Regularly Varying Values ...... 3-6
3.8 Trace Header Manipulation 2: Reverse Numbering ................................... 3-8
3.9 Windowing Data ......................................................................................... 3-10
4. Three Simple Models .......................................................................................
4.1 Introduction ................................................................................................ 4-1
4.2 Model 1: Earth Model – Five Layers ......................................................... 4-1
4.3 Model 2: Earth Model – High Velocity Intrusion ...................................... 4-5
4.4 Model 3: Earth Model – Diffractor ............................................................ 4-7
4.5 Introduction to Ray Tracing and Wavefronts ............................................. 4-9
4.6 Model 1 Ray Tracing: psmerge1a.sh ......................................................... 4-10
4.7 Model 1 Wavefronts: psmerge1b.sh ........................................................... 4-11
4.8 Model 2 Ray Tracing: psmerge2a.sh ......................................................... 4-13
4.9 Model 2 Wavefronts: psmerge2b.sh ........................................................... 4-14
4.10 Model 3 Ray Tracing: psmerge3.sh ........................................................... 4-16
5. Three Simple Models: Acquire 2-D Lines .......................................................
5.1 Introduction ................................................................................................ 5-1
5.2 Acquire 2-D Line from Model 1: acq1.sh .................................................. 5-1
5.3 Acquire 2-D Line from Model 2: acq2.sh .................................................. 5-5
5.4 Acquire 2-D Line from Model 3: acq3.sh .................................................. 5-5
5.5 Computer Note 7: Pause and Resume Screen Print ................................... 5-6
5.6 Examine Shot Gathers: showshot.sh .......................................................... 5-6

vi
TOC-2
 Table of Contents

5.7 A Graphics Program is a Terminal Program .............................................. 5-8

5.8 Computer Note 8: Use, Modify a Command from the History List ........... 5-8
5.8.1 Use a Command from the History List ................................................ 5-8
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

5.8.1.1 Up Arrow Key ................................................................................ 5-8

5.8.1.2 History List ..................................................................................... 5-8
5.8.2 Modify a Command from the History List ........................................... 5-8
5.9 Survey Quality Control .............................................................................. 5-9
6. Model 4: Build, Acquire a Line, Display Gathers, QC ....................................
6.1 Introduction ................................................................................................ 6-1
6.2 Model 4: Earth Model – Diffractor and Pinchout ...................................... 6-1
6.3 Acquire 2-D Line from Model 4: acq4.sh .................................................. 6-5
6.4 Model-Acquisition Checklist ..................................................................... 6-9
6.5 Examine Shot Gathers ................................................................................ 6-10
6.6 Computer Note 9: Data Files Not in Script Directory ................................ 6-14
6.6.1 Data Directory below Script Directory — Full Path Name ................. 6-14
6.6.2 Data Directory below Script Directory — Short Path Name ............... 6-15
6.6.3 Data Directory at the same Level as Script Directory .......................... 6-16
6.7 Quality Control: The Survey File ............................................................... 6-17
6.8 Quality Control: Acquire and Display Specific Shot Gathers .................... 6-18
7. Model 4: Sort, Velocity Analysis .....................................................................
7.1 Introduction ................................................................................................ 7-1
7.2 Sort from Shot Gathers to CMP Gathers .................................................... 7-1
7.3 Extract CMP Gathers .................................................................................. 7-3
7.4 Computer Note 10: If Blocks ..................................................................... 7-4
7.5 Fold ............................................................................................................. 7-5
7.6 Stacking Velocity Analysis ........................................................................ 7-5
7.6.1 Prepare a Shot Gather for Velocity Analysis ....................................... 7-5
7.6.2 Semblance Contour Plot ....................................................................... 7-8
7.6.3 Constant Velocity Stacks ...................................................................... 7-11
7.6.4 Quality Control (QC) after Picking ...................................................... 7-12
7.6.5 Computer Note 11: Interactive Scripts ................................................. 7-13
7.6.6 Velocity Analysis: Scripts iva.scr and iva.sh ....................................... 7-13
7.6.6.1 iva.scr ............................................................................................. 7-14
7.6.6.2 iva Displays .................................................................................... 7-14
7.6.6.3 iva.sh — Introduction ..................................................................... 7-20
7.6.6.4 iva.sh — User-supplied Values ...................................................... 7-28
7.6.6.5 iva.sh — File Documentation ......................................................... 7-29
7.6.6.6 iva.sh — One-time Processing ....................................................... 7-30

vii
TOC-3
 Table of Contents

7.6.6.7 iva.sh — IVA Loop ........................................................................ 7-30

7.6.6.8 iva.sh — Output the Picks .............................................................. 7-32
7.6.6.9 iva.sh — Pause, Remove Temporary Files, Exit ............................ 7-33
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

7.6.6.10 iva.sh — vpick.txt ........................................................................... 7-33

7.7 X-axis Labels in suxwigb and suximage ................................................... 7-33
8. Model 4: T-V Picks QC, NMO, Stack .............................................................
8.1 Introduction ................................................................................................ 8-1
8.2 Time-Velocity QC ...................................................................................... 8-1
8.2.1 Program tvnmoqc, mode=1 ................................................................. 8-1
8.2.2 Program tvnmoqc, mode=2 ................................................................. 8-2
8.2.2.1 Script iva.scr ................................................................................... 8-5
8.2.2.2 Script iva.sh .................................................................................... 8-5
8.2.3 Edit Time Picks .................................................................................... 8-9
8.3 Normal Moveout: Script nmo4.sh .............................................................. 8-10
8.4 Interactive Viewer ...................................................................................... 8-14
8.4.1 Script iview.scr ..................................................................................... 8-14
8.4.2 Script iview.sh ...................................................................................... 8-14
8.5 Stack ........................................................................................................... 8-17
9. Model 4: Migration ..........................................................................................
9.1 Introduction ................................................................................................ 9-1
9.2 Migration Constant Velocity Panels (CVP) with Stolt Migration ............. 9-1
9.2.1 sustolt Parameters ................................................................................ 9-1
9.2.2 Seismic Migration with Model 4 and sustolt ....................................... 9-2
9.2.3 Migration Constant Velocity Panels (CVPs) ........................................ 9-3
9.3 Interactive Constant Velocity Panel Viewer .............................................. 9-6
9.3.1 Script iviewcvp.scr ................................................................................ 9-6
9.3.2 Script iviewcvp.sh ................................................................................. 9-6
9.4 Migration Movie ......................................................................................... 9-11
9.5 Migrate Model 4 ......................................................................................... 9-12
10. Nankai Data: Examine, Resample, Sort, Gain .................................................
10.1 Introduction ................................................................................................ 10-1
10.1.1 Nankai Shot Gathers ............................................................................. 10-1
10.1.2 Nankai Stack File ................................................................................. 10-2
10.1.3 Our Task ............................................................................................... 10-3
10.2 Computer Note 12: Symbolic Link ............................................................ 10-4
10.3 View the Line of Shot Gathers ................................................................... 10-5
10.4 f-x Display, Filter, Resample, and Sort Nankai Data ................................. 10-7
9.4.1 Script fxdisp .......................................................................................... 10-8

viii
TOC-4
 Table of Contents

9.4.2 Anti-alias Filter, Resample, and Sort ................................................... 10-11

10.5 Gain -- Spherical Divergence Correction ................................................... 10-13
10.6 Single-trace versus Multi-trace Processes .................................................. 10-21
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

10.7 Summary .................................................................................................... 10-21

11. Nankai: Velocity Analysis, NMO, Stack, Migration .......................................
11.1 Velocity Analysis ....................................................................................... 11-1
11.2 NMO and Stack .......................................................................................... 11-3
11.3 Stolt Migration with sustolt ....................................................................... 11-6
11.3.1 Script Nmigcvp.sh ................................................................................. 11-6
11.3.2 Script Niviewcvp ................................................................................... 11-6
11.3.3 Script Nmigmovie.sh ............................................................................. 11-7
11.3.4 Summary .............................................................................................. 11-8
11.4 Phase Shift Migration with sugazmig ........................................................ 11-8
11.5 Some Geologic Features ............................................................................. 11-10
11.5.1 Bottom-simulating Reflector ................................................................ 11-10
11.5.2 Subducted Pacific Plate ........................................................................ 11-10
12. Taiwan Data: Examine, Zero Traces, Re-examine ...........................................
12.1 Introduction ................................................................................................ 12-1
12.2 Reminder: Symbolic Link .......................................................................... 12-2
12.3 View the Line of Shot Gathers ................................................................... 12-2
12.4 Kill Traces .................................................................................................. 12-5
12.5 One-dimensional Frequency Analysis ........................................................ 12-10
13. Taiwan: Gain-Filter, Filter-Gain ......................................................................
13.1 Introduction ................................................................................................ 13-1
13.2 Gain-Filter .................................................................................................. 13-1
13.2.1 Gain – Spherical Divergence Correction .............................................. 13-1
13.2.2 Band-pass Filter .................................................................................... 13-3
13.2.3 Gain-Filter Result ................................................................................. 13-5
13.3 Filter-Gain .................................................................................................. 13-6
13.3.1 Band-pass Filter .................................................................................... 13-6
13.3.2 Gain – Spherical Divergence Correction .............................................. 13-6
13.3.3 Filter-Gain Result ................................................................................. 13-9
13.4 Discussion .................................................................................................. 13-10
14. f-k Filter and Deconvolution ............................................................................
14.1 Introduction ................................................................................................ 14-1
14.2 f-k Filter – Absolute Units .......................................................................... 14-1
14.2.1 Use the f-k Filter Script ........................................................................ 14-1
14.2.2 Script ifk.scr .......................................................................................... 14-3

ix
TOC-5
 Table of Contents

14.2.3 Script ifk.sh ........................................................................................... 14-5

14.3 f-k Filter – Relative Units ........................................................................... 14-10
14.4 f-k Filter – Discussion ................................................................................ 14-14
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

14.5 Deconvolution ............................................................................................ 14-18

14.5.1 Use Deconvolution Script idecon ......................................................... 14-18
14.5.2 Script idecon.scr ................................................................................... 14-24
14.5.3 Script idecon.sh .................................................................................... 14-24
15. Taiwan: Sort, Velocity Analysis, NMO, Stack ................................................
15.1 Introduction ................................................................................................ 15-1
15.2 Sort ............................................................................................................. 15-1
15.3 Velocity Analysis ....................................................................................... 15-2
15.4 NMO and Stack .......................................................................................... 15-4
16. Taiwan: Migration ............................................................................................
16.1 Constant Velocity Stolt Migration with Tmigcvp.sh .................................. 16-1
16.2 Migration Movie ......................................................................................... 16-1
16.2 Discussion .................................................................................................. 16-3
References ................................................................................................................
Appendix A. Seismic Un*x Legal Statement ........................................................
Appendix B. Seismic Un*x at Michigan Tech ......................................................
B.1 Using Seismic Un*x ................................................................................... B-1
B.2 Terminal Windows ..................................................................................... B-1
Appendix C. Utility Programs ...............................................................................
C.1 Introduction ................................................................................................ C-1
C.2 Program sukeycount .................................................................................. C-1
C.3 Program sudumptrace ............................................................................... C-5
C.4 Program tvnmoqc ...................................................................................... C-12
C.5 Conclusion .................................................................................................. C-15
Appendix D. Makefiles: Alternative to Shell Scripts, by Chris Liner ...................
D.1 Introduction ................................................................................................ D-1
D.2 Discuss the Makefile .................................................................................. D-1
D.3 “clean” and Final Thoughts ........................................................................ D-4
D.4 The Makefile .............................................................................................. D-4
D.5 The Makefile Edited ................................................................................... D-5
Appendix E. On the CDs .......................................................................................
E.1 Category List .............................................................................................. E-1
E.2 Alphabetical List – Unix (excluding figures) ............................................. E-3
E.3 Alphabetical List – PC (excluding figures) ................................................ E-3

x
TOC-6
 Preface
This book can serve either of two purposes. (1) It can be used, as it is in our courses
at Michigan Technological University, as an aid to teaching seismic reflection data
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

processing. (2) It can be used as a primer to Seismic Un*x (SU) by those who may or
may not already be familiar with seismic processing using other software packages. SU is
provided by the Center for Wave Phenomena at Colorado School of Mines and is
available from their web site www.cwp.mines.edu/cwpcodes.
There are details of SU that are important to the processing specialist, but are not
essential to the student who is being introduced to seismic processing. Where these
details appear in the text, we placed a gray bar in the margin to indicate that the material
can be skipped by the student interested only in learning about processing. Of course,
these details might be among the most-interesting material from the viewpoint of
someone who wants to learn programming techniques that apply to SU.
A beginning course in reflection seismology processing might complete the entire
book in one semester. A course that covers fundamental processing and some
interpretation might skip the material with a gray bar and use only one of the real (field,
not synthetic) data sets that we provide.
We have found that one of the biggest hurdles to developing a new course is the
scarcity of real data. With this book, we provide two real data sets. The first real data set,
from the Nankai trough near Japan, is unusual in that it was acquired over very deep
water. This presents some advantages (mediocre velocity analyses will still produce good
results), but it is not realistic for typical exploration purposes. It is, however, a site of
exciting geologic features, and the rugged seafloor topography dramatically demonstrates
the benefit of migration. It is a data set that any student can appreciate. The second real
data set, from offshore Taiwan, presents a number of processing challenges; it is much
more difficult to process to satisfaction.
The compact disks (CDs) included with this book have copies of the scripts and the
seismic data sets (including some data sets that are generated by the scripts). The book is
printed in black-and-white to keep its price low. Because some figures are best viewed in
color, we put copies of all color figures, in uncompressed TIFF format, on the CDs.
Appendix E has a complete list of the contents of the CDs.
We at Michigan Tech will maintain a list of errata for the book. The errata list can be
found by searching on the web for Michigan Tech, errata, and Seismic Un*x. Please
report any errors to the address identified on that web site. Reports of SU bugs,
suggestions for SU improvements, or proposals for new SU scripts should be submitted
to the CSM Center for Wave Phenomena via methods suggested on their web site.

xi
P-1
 Table of Computer Notes
Computer Notes are guides to help you better understand and use the Unix system.
Chapter 2 is not listed here, but it can be considered a chapter of Computer Notes.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Warning 1: We developed this Seismic Un*x Primer while using the csh shell. The
Computer Notes were written from this perspective. If you are not working under the csh
shell, our Computer Notes might not be appropriate for you.
Warning 2: The previous warning does not apply to the scripts. The first line of every
script is, “#! /bin/sh” , a command that makes scripts run under the Bourne shell.

Section Page
1.4 Computer Note 1: Enter ............................................................................. 1-2
1.5 Computer Note 2: Your Shell ..................................................................... 1-3
1.6 Computer Note 3: Changing Directories .................................................... 1-3
1.7 Computer Note 4: Cursor Prompt .............................................................. 1-3
1.9 Computer Note 5: Background (&) and Foreground Processes ................. 1-5
1.9.1 Background Processes .......................................................................... 1-5
1.9.2 Foreground Processes ........................................................................... 1-6
1.13.1 Computer Note 6: No Spaces around the Equal Sign .......................... 1-9
5.5 Computer Note 7: Pause and Resume Screen Print ................................... 5-6
5.8 Computer Note 8: Use, Modify a Command from the History List ........... 5-8
5.8.1 Use a Command from the History List ................................................ 5-8
5.8.2 Modify a Command from the History List ........................................... 5-8
6.6 Computer Note 9: Data Files Not in Script Directory ................................ 6-14
6.6.1 Data Directory below Script Directory — Full Path Name ................. 6-14
6.6.2 Data Directory below Script Directory — Short Path Name ............... 6-15
6.6.3 Data Directory at the same Level as Script Directory .......................... 6-16
7.4 Computer Note 10: If Blocks ..................................................................... 7-4
7.6.5 Computer Note 11: Interactive Scripts ................................................. 7-13
10.2 Computer Note 12: Symbolic Link ............................................................ 10-4

Throughout this Primer, we use a vertical bar, as shown to the right, to mark passages
that a first-time reader can safely ignore. Generally, the bar marks two kinds of text.
Details of model building (large portions of Chapters 4, 5, and 6) can be ignored because
Model 4 is on one of the CDs. The bar also marks parts of scripts and explanations of
those parts that are of more interest to shell script programmers than to script users.

xii
TOCN-1
 Table of Scripts
Scripts are listed when introduced. The page number is the page of the first line of the
script. When a script “fragment” is listed, the original script location is also cited.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Line commands are not referenced, with the exception of our only use of sumute.

Section Page
2.4 myplot.sh ...................................................................................................... 2-5
4.2 model1.sh ..................................................................................................... 4-1
4.3 model2.sh ..................................................................................................... 4-5
4.4 model3.sh ..................................................................................................... 4-7
4.6 psmerge1a.sh ................................................................................................ 4-10
4.7 psmerge1b.sh ............................................................................................... 4-12
4.8 psmerge2a.sh ................................................................................................ 4-13
4.9 psmerge2b.sh ............................................................................................... 4-15
4.10 psmerge3.sh ................................................................................................. 4-17
5.2 acq1.sh ......................................................................................................... 5-1
5.3 acq2.sh (fragment, modification of acq1.sh, 5.2) ........................................ 5-5
5.4 acq3.sh (fragment, modification of acq1.sh, 5.2) ........................................ 5-5
5.6 showshot.sh .................................................................................................. 5-6
6.2 model4.sh ..................................................................................................... 6-1
6.3 acq4.sh ......................................................................................................... 6-5
6.5 showshot.sh (repeated from 5.6) .................................................................. 6-10
6.6.1 shotrwd1.sh ............................................................................................... 6-14
6.6.2 shotrwd2.sh ............................................................................................... 6-15
6.6.2 shotrwd3.sh ............................................................................................... 6-16
6.6.3 shotrwd4.sh ............................................................................................... 6-16
6.8 acq4shot.sh (fragment, modification of acq4.sh, 6.3) .................................. 6-18
6.8 showshotB.sh ............................................................................................... 6-18
7.2 sort2cmp.sh .................................................................................................. 7-1
7.3 showcmp.sh .................................................................................................. 7-4
7.6.1 oz14prep.sh ............................................................................................... 7-6
7.6.2 oz14velan.sh .............................................................................................. 7-9
7.6.5 iva.scr ........................................................................................................ 7-13
7.6.6.1 iva.scr (repeated from 7.6.5) .................................................................. 7-14
7.6.6.3 iva.sh ...................................................................................................... 7-21
8.2.1 tvQC.sh ..................................................................................................... 8-1
8.2.2.1 velanQC.scr ............................................................................................ 8-5

xiii
TOS-1
 Table of Scripts

8.2.2.2 velanQC.sh ............................................................................................. 8-5

8.3 nmo4.sh ........................................................................................................ 8-11
8.3 nmo4a.sh ...................................................................................................... 8-13
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

8.4.1 iview.scr .................................................................................................... 8-14

8.4.2 iview.sh ..................................................................................................... 8-14
9.2.3 migcvp.sh .................................................................................................. 9-4
9.3.1 iviewcvp.scr .............................................................................................. 9-6
9.3.2 iviewcvp.sh ............................................................................................... 9-7
9.4 migmovie.sh ................................................................................................. 9-11
9.5 migStolt.sh ................................................................................................... 9-12
10.3 iview.sh (fragment, modification of 8.3.2) .................................................. 10-5
10.4.1 fxdisp.sh .................................................................................................... 10-8
10.4.1 fxdisp.scr ................................................................................................... 10-9
10.4.2 Nsort2cmp.sh (includes sufilter, suresamp) ............................................ 10-11
10.5 igain.scr ........................................................................................................ 10-13
10.5 igain.sh ......................................................................................................... 10-13
11.1 iva.sh (fragment, modification of 7.6.6.3) ................................................... 11-1
11.2 Nnmo.sh ....................................................................................................... 11-4
11.3.1 Nmigcvp.sh (fragment, modification of migcvp.sh, 8.5.3) .................. 11-6
11.3.2 Niviewcvp.sh (fragment, modification of iviewcvp.sh, 8.6.2) ............. 11-6
11.3.3 Nmigmovie.sh ........................................................................................... 11-7
11.4 Ngazmig.sh .................................................................................................. 11-9
12.3 iview.sh (fragment, modification of 8.3.2) .................................................. 12-2
12.4 killer.sh ......................................................................................................... 12-5
12.4 killer2.sh ....................................................................................................... 12-6
12.5 tf.sh ............................................................................................................... 12-11
13.2.1 igain.sh (fragment, modification of 9.5) ............................................... 13-1
14.2.2 ifk.scr ........................................................................................................ 14-3
14.2.3 ifk.sh .......................................................................................................... 14-5
14.5.1 sumute (line command example) ............................................................. 14-19
14.5.2 idecon.scr .................................................................................................. 14-24
14.5.3 idecon.sh ................................................................................................... 14-24
15.3 iva.sh (fragment, modification of 7.6.6.3) ................................................... 15-2
15.4 Tnmo.sh ....................................................................................................... 15-4
16.1 Tmigcvp.sh (fragment, modification of migcvp.sh, 8.5.3) .......................... 16-1
16.2 Tmigmovie.sh .............................................................................................. 16-1

TOS-2
 Introduction

1. Introduction
1.1 The Goal of this Primer
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Our objective is to introduce you to the fundamentals of seismic data processing with
a learn-by-doing approach. We do this with Seismic Un*x (SU), a free software package
maintained and distributed by the Center for Wave Phenomena (CWP) at the Colorado
School of Mines (CSM). At the outset, we want to express our gratitude to John
Stockwell of the CWP for his expert counsel.
SU runs on several operating systems, including Unix, Microsoft Windows, and
Apple Macintosh. However, we discuss SU only on Unix.
Detailed discussion of wave propagation, convolution, cross- and auto-correlation,
Fourier transforms, semblance, and migration are too advanced for this Primer. Instead,
we suggest you refer to other publications of the Society of Exploration Geophysicists,
such as “Digital Processing of Geophysical Data - A Review” by Roy O. Lindseth and
one of the two books by Ozdogan Yilmaz: “Seismic Data Processing,” 1987 and
“Seismic Data Analysis,” 2001.
Our goal is to give you the experience and tools to continue exploring the concepts of
seismic data processing on your own.

1.2 The Outline of this Primer

This Primer covers all processing steps necessary to produce a time migrated section
from a 2-D seismic line. We use three sources of input data:
ƒ Synthetic data generated by SU;
ƒ Real shot gathers from the Oz Yilmaz collection at the Colorado School of Mines
(ftp://ftp.cwp.mines.edu/pub/data); and
ƒ Real 2-D marine lines provided courtesy of Prof. Greg Moore of the University of
Hawaii: the “Nankai” data set and the “Taiwan” data set.
o The University of Texas, the University of Tulsa, and the University of Tokyo
collected the Nankai data. The U.S. National Science Foundation and the
government of Japan funded acquisition of the Nankai data.
o The University of Hawaii, San Jose State University, and National Taiwan
University collected the Taiwan data. The U.S. National Science Foundation
and the National Science Council of Taiwan funded acquisition of the Taiwan
data.
Chapters 1-3 introduce the Unix system and Seismic Un*x.
Chapters 4-5 build three simple models (complexity slowly increases) and acquire a
2-D line over each model. (These chapters may be skipped if you are only interested in
processing.)
Chapters 6-9 build a model based on the previous three, acquire a 2-D line over that
model, and process the line through migration.
Chapters 10-11 start with a real 2-D seismic line of shot gathers (Nankai) and process
it through migration.

1-1
 Introduction

Chapters 12-13 and 15-16 start with a real 2-D line of shot gathers (Taiwan) and
process it through migration.
Chapter 14 uses real shot gathers from the Oz Yilmaz collection to demonstrate f-k
filtering and deconvolution.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

We introduce the major processing steps in the following chapters:

Build a model Chapter 4
Acquire seismic data Chapter 5
Amplitude correction Chapter 10
Frequency (1-D) filtering Chapter 10
Frequency-wavenumber (f-k) filtering Chapter 14
Deconvolution Chapter 14
Sort from shot gathers to CMP gathers Chapter 7
Velocity analysis Chapter 7
Normal moveout correction (NMO) Chapter 8
Stack Chapter 8
Migration Chapter 9
Scripts are included for each process. A detailed description of these scripts is the
heart of this Primer.
Experience with the Unix operating system is helpful, but not necessary. Chapter 2
introduces Unix commands and SU scripts.

1.3 A Bit about Seismic Un*x

Seismic Un*x (Cohen & Stockwell, 2002) is a free software package for seismic data
processing. SU is maintained and regularly updated by the Center for Wave Phenomena
(CWP) at the Colorado School of Mines (CSM). The SU home page at the CWP is:
http://www.cwp.mines.edu/cwpcodes/
You can download SU from the CWP ftp server and install it on almost any Unix system.
The source code is also included, making it possible for you to adjust and expand SU.
The CWP web site also contains links to other useful web sites.
Please read the legal statement that accompanies SU. For your convenience, we make
this available as our Appendix A.
SU is easy to use because it does not require extensive programming knowledge, only
basic Unix commands for file management, pipes, and shell redirection (see Chapter 2).
Although SU can be run from the command line, the most efficient way to use SU is
through shell scripts. Shell scripts let us combine several processing programs into one
"job," similar to commercial seismic processing packages. We first describe Shell scripts
in Chapter 2.

1.4 Computer Note 1: Enter

When we tell you to “enter” something on the computer, we want you to type the
instruction, then press the Enter or the Return key. Pressing Enter or Return tells the
computer to execute the instruction. Merely typing an instruction on a line does not
execute the instruction.

1-2
 Introduction

1.5 Computer Note 2: Your Shell

When you enter a command, you are interacting with the Unix shell. There are
several Unix shells: the Bourne shell, the C shell, the Korn shell, the Bourne Again shell
(BASH), and others. All example scripts in this Primer use the Bourne shell; most of
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

the Computer Notes refer to the C shell.

1.6 Computer Note 3: Changing Directories

1. Open an x-term window and enter pwd (on some computers you have to use cwd)
to see the directory in which you are now working.
2. Enter ls to see a listing of the files and subdirectories in that directory.
3. If you have a subdirectory; for example, “work,” enter cd work to change to
that subdirectory, then enter ls to see a listing within it.
4. When you want to move up one level of directory, enter cd .. (Yes, that is: cd
space dot dot.)
5. If you enter cd by itself, you will go to your home directory.
6. If you enter cd . (cd space dot), you will not go anywhere because the
single dot (period) is a reference to your current directory.
The single dot can be useful. Suppose there is a system calculator called xcalc
and, suppose you wrote your own calculator program that is also called xcalc. If
you enter
xcalc
the system version might start instead of yours. To run your version instead of the
system version, go to the directory that contains your version, then enter
./xcalc
The dot slash tells the system to look in the current directory for the program.

1.7 Computer Note 4: Cursor Prompt

Usually, the default computer prompt does not give much information. For example,
Figure 1.1 shows the default prompt when I (Forel) open a new x-term window

Figure 1.1: Forel’s default x-term prompt.

The prompt of Figure 1.1 shows the name of the machine that I am on and the number
of the command (1) that I am about to enter. I think a more useful prompt is one that tells
me which directory I am in (Figure 1.2).
A disadvantage of a prompt that shows the current directory is that as I go to deeper
subdirectories, the prompt becomes longer (Figure 1.3). I am willing to pay this price for
the information that I gain.

1-3
 Introduction

/home/forel>
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 1.2: My new prompt shows the current directory I also changed “%’ to “>”.

/home/forel/seismicx/demos/Filtering/Sufilter>

Figure 1.3: A prompt with a long path name.

To change your C shell prompt, edit your home directory’s shell resource file. The
file “.cshrc” is the resource file for the C shell. Yes, the file name starts with a period (.).
Use any editor you like to add the following lines to your “.cshrc” file.
Note: We use only Unix-based text editors to write and edit shell scripts and resource
files. Many personal computer editors insert hidden format codes that might be wrongly
interpreted by the Unix operating system.
# Make prompt contain "pwd"
alias setprompt 'set prompt="$cwd> "'
alias cd 'cd \!* && setprompt'
alias pushd 'pushd \!* && setprompt'
alias popd 'popd \!* && setprompt'
setprompt
The first line is a comment. When the first character is “#,” the entire line is ignored.
After you save your changes, when you open a new x-term window, the system will
read the resource file and use those instructions in the new x-term window. If you want to
use .cshrc changes in your current window, enter:
source .cshrc

1.8 Your First Seismic Un*x Instruction

Before you can run Seismic Un*x, you must be sure your SU commands will be
interpreted by the Unix operating system. This means SU must be “in your path.” Your
system administrator may have made SU part of your login path or you may have to enter
a command first. The approach we use at Michigan Tech is described in Appendix B.
Let's run a simple program to test that you and your SU environment are working
together. Enter (type the following, then press the Enter or the Return key):
suplane | suxwigb &
You should get a new window with a picture of some synthetic seismic data on your
screen (Figure 1.4). If you do not, ask someone for help.
The style of the above command is:
[program] [“pipe” output data to next program] [program] &
The pipe “|” directs data from one process to another.

1-4
 Introduction

You can regain the Unix prompt in your SeismicX window by pressing the return key
(if you supplied & at the end of the command). You can close the graphics window by
using the Unix Close command (the top left button).
Let’s discuss the details of the above command:
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

• suplane creates seismic data.

• the pipe (the vertical bar) directs (sends) the data to the next program.
• suxwigb (the X-window display program) displays the seismic data on your
screen as wiggle traces.
• The & needs a special explanation. Read the next section.

Figure 1.4: The output of your first SU command. The upper-left corner shows
(0.200719,8.05814) because we clicked the middle mouse button at approximately 0.200
ms time (the vertical axis is time) and on the eighth trace from the left (the horizontal axis
is trace number).

1.9 Computer Note 5: Background (&) and Foreground Processes

1.9.1 Background Processes

In technical language, using & (ampersand) at the end of a command runs (spawns)
the process in the background. In simple terms, using & does not lock your cursor. Since
your cursor is not locked (frozen) when you use & at the end of a command, you can use
the same x-term to do other things, like submit other processes. For example, the
following command starts the Netscape web browser as a background process:
netscape &
To enter other commands after this one, press Enter again to regain a prompt, then
enter your next command.

1-5
 Introduction

While a process is running in the background, you might need to cancel (kill) it. You
can do this using a Unix utility called top.
1. In an x-term window, enter top
2. The x-term window now shows your most active processes (Figure 1.5) and the
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

cursor is now in the upper left of the top display (shown as the red box in Figure
1.5).

Figure 1.5: Screen capture of Forel’s top processes.

3. The left column (PID) identifies process netscape, the Netscape web browser. as
9828. To kill this process, enter
k 9828
4. To exit top, enter q for quit.
Note: You can use top to learn the process ID (PID), then kill the process using a line
command. To kill the Netscape process described above using a line command, enter:
kill -9 9828
The flag “-9” tells kill that you are serious about killing process 9828.

1.9.2 Foreground Processes

When you start a process and don't use &, the new process uses the current x-term
(the process runs in the foreground), which locks the cursor. For example, the following
command starts the Netscape web browser as a foreground process:
netscape
To continue working when the cursor is locked, you must either kill the process that
you just started or you must open a new x-term. To kill a foreground process in an x-term
window:
1. click the window to make it active,
2. hold down the Control key and briefly press the c key (press Control-c).
This kills the process. Because the process dies, the browser window also closes.

1.10 Zoom the xwigb Window

You can zoom the image in the xwigb window. Put the cursor somewhere in the
xwigb window, press and hold down the left mouse button, drag the mouse until the box
encloses the area you want enlarged, then let go of the button.
You can zoom repeatedly. Notice that time numbers and trace numbers correspond to
the zoomed data. Click once in the window to return to the original view.

1-6
 Introduction
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 1.6: Zoom an xwigb window by dragging a box with the left mouse button. The
left image shows the mouse dragging the box. The right image shows the zoomed image.
To return to the original view, click the left mouse button in the window.

1.11 The Demos Directory

The demos directory is a standard part of the SU installation. If you don’t know
where yours is, ask someone. It is full of easy to use demonstrations of many seismic data
processing programs.
Once you are in the demos directory, repeatedly use cd to move around. Use ls and
pwd frequently to see where you are. Notice the many README files. The README
files tell how to use the demonstrations in each directory.You can write README files
to your screen by typing either of the following commands:
cat README
more README
If you use the second command, you can advance the text that appears on your screen
by pressing the spacebar to move to the next page. Pressing Enter or Return scrolls the
display one line.
Exploring (executing) the demonstrations is an easy way to see seismic data
processing!

1.12 Self Documentation in Seismic Un*x

To display documentation for programs like suplane or suxwigb, enter the name of
the program without arguments. "Without arguments" means you enter only the name of
an SU program. For example, enter
suplane

1-7
 Introduction

SU interprets this to mean you want information about suplane, so SU prints the program
help (selfdoc) file to the screen. Below are the first several lines of the suplane selfdoc
file.
SUPLANE - create common offset data file with up to 3 planes
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

suplane [optional parameters] >stdout

Optional Parameters:
npl=3 number of planes
nt=64 number of time samples
ntr=32 number of traces
taper=0 no end-of-plane taper
= 1 taper planes to zero at the end
offset=400 offset
dt=0.004 time sample interval in seconds
You see that the name stands for “create common offset data file with up to 3 planes.”
(Many SU program names start with “su.”) You also see the default values (values used
when the user does not supply a value) of various parameters, such as the number of
planes (npl=3), the number of time samples per trace (nt=64) and the time sample
interval (dt=4 ms). Using default values, each trace is 256 ms long.

1.13 Your Second Seismic Un*x Instruction

Let’s change the command given earlier in Section 1.8. Enter:
suplane npl=2 dt=0.008 | suxwigb &
By supplying values for npl and dt, you overwrite (replace) suplane’s default values.
Your output (pictured below) has only two planes and each trace is 512 ms long.

Figure 1.7: The suplane command with two defaults overwritten: number of planes = 2,
and time interval = 0.008 seconds.

1-8
 Introduction

1.13.1 Computer Note 6: No Spaces around the Equal Sign

Do not use spaces before or after the equal (=) sign

1.14 Other Help Facilities in Seismic Un*x

Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Various help facilities are described in The New SU User’s Manual by John W.
Stockwell, Jr. & Jack K. Cohen; Version 3.2: August 2002. This Manual is available at
the CWP Seismic Un*x web site. In addition to the self documentation described in
Section 1.11, other help facilities include:
• suhelp lists all the available programs.
• suname lists all programs and libraries with a short description of them.
• sudoc followed by the program name gives documentation of the program. This
can work even when there is no selfdoc for the program.
• sufind followed by a string searches in all self documentation for this string.
• sukeyword lists keys used for headers.
• demos is a directory in the SU installation that contains useful scripts.

1-9
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

This page has been intentionally left blank

 Unix Commands and Concepts

2. Unix Commands and Concepts

2.1 Elementary Unix Commands
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

The following table lists essential Unix commands.

Table 2.1: Unix commands
Command Description
| The “pipe” connects two processes.
& Run (spawn) a process in the background.
cat Display a text file on the screen. Example: cat zebra.txt
cd Change to home directory or change to directory specified after
space.
cd .. Go up one directory
chmod Change file permissions mode. Example: chmod +x filter.su
cp Make a copy of a file. Example: cp zebra2.txt zebra3.txt
find Search for (find) files within and below a directory.
Example: find . -name ’acq*.sh’ –print
ls List directories and files.
ls -a List directories and files, including hidden files (files with a dot
prefix).
ls –lF List directories and files using long format.
man Access Unix (not Seismic Un*x) help (manual) pages.
Example: man ls
mkdir Make directory. Example: mkdir data
more Display the contents of a text file, one screen at a time. If the file
occupies more than one screen, press the space bar to scroll a page
or the Return (Enter) key to scroll a line.
mv Move or rename a file or directory.
Example: mv zebra2.txt zebra3.txt
pwd (or cwd) Show “present working directory.”
rm Remove (delete) a file. Example: rm zebra.txt
rmdir Remove a directory. (A directory must usually be empty before it
can be removed.) Example: rmdir data
Control-c Kill (stop) a process using the active window.

2-1
 Unix Commands and Concepts

This chapter explains fundamental Unix commands that are necessary for
understanding later scripts. It will be helpful if you know elementary Unix commands.
Books titled “Teach Yourself Unix” have excellent, simple, early chapters that give the
basics. Also, by surfing the web, you can find universities that have good tutorial sites.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

2.2 File Name Conventions

The following file name suffixes are used throughout this Primer.
• .sh shell script
• .scr shell script that launches a .sh shell script
• .su binary seismic data
• .dat binary data, not seismic data
• .eps image file formatted as Encapsulated Postscript (EPS)
• .txt ASCII data file
The SU seismic data format is based on, but is not identical to, the binary format
called SEG-Y. SEG-Y was defined by the Society of Exploration Geophysicists (SEG)
and has become an industry standard format for seismic data exchange.
The following file types can be printed directly to the screen by the cat and more
commands: .sh, .scr, .txt.

2.3 Advanced Unix Concepts

We do not explain the following advanced Unix concepts in detail. Their usage will
be made clear by the way we use them in later scripts.
The following Unix commands are not complete. These are merely a selection of
commands that we consider helpful for your understanding and reproduction of the
processing in this Primer.
Remember that Unix is case sensitive. That is, suplane is not the same as Suplane.

2.3.1 Invoke the Shell: #! /bin/sh

To start the Bourne shell interpreter, the first line of any script we make must be:
#! /bin/sh

2.3.2 Comment Line: #

A comment line is one that is not seen by the shell. It is for user-comments within a
shell script. A comment line begins with the # character. For example:
# Name input and output files
Note: The use of “#” shown in the previous section is a rare exception to the use of “#”
explained here.
Note: The shell ignores blank lines.
Note: Do not insert comment lines or blank lines in the middle of an instruction.

2.3.3 Pipe: |
We saw the pipe in Section 1.8. We mention it here because it is an advanced
concept. A pipe allows data to flow from one process to another. Below (as before),

2-2
 Unix Commands and Concepts

suplane creates data, then the pipe sends the data to the imaging program, suxwigb to be
seen on the screen. In other words, data flow is through the pipe, left to right.
suplane | suxwigb
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

2.3.4 Shell Redirection: < and > and >>

The redirect < sends data from right to left. Below, data file seis1.su is sent as input to
program suwind, a windowing program that we will use later.
suwind < seis1.su
The redirect combination below sends data file (a) to process (b); the output of (b) is
stored in file (c).
suwind < seis1.su > seis2.su
b < a > c
The first command below types the contents of file README to the screen. The
second command below redirects the result of the cat command to file info2.txt; the
contents of file README do not get to the screen. The third command below appends the
contents of file README to the bottom of file info2.txt; that is, the original contents of
file info2.txt are not overwritten.
cat README
cat README > info2.txt
cat README >> info2.txt

2.3.5 Permissions: rwxrwxrwx

All files on a Unix system have permission rights associated with them. These rights
of “read” (r), “write” (w), and “execute” (x) are assigned to “user,” “group,” and “other.”
You can view the permissions of a file by entering ls –l. (We prefer ls –lF.) The
file permissions are in the left column. For example:
/home/forel/suscripts> ls -lF
total 160
-rw-r--r-- 1 forel geograd 1915 Apr 15 19:02 nmoall.sh
drwxr-xr-x 2 forel geograd 512 Apr 15 11:08 data/
Permissions can be changed with the command chmod (change the permissions mode
of a file). SCRIPT FILES MUST BE EXECUTABLE. To make a file executable, enter
chmod +x and the file name. For example:
/home/forel/suscripts> chmod +x nmoall.sh
/home/forel/suscripts> ls -lF
total 160
-rwxr-xr-x 1 forel geograd 1915 Apr 15 19:02 nmoall.sh*
drwxr-xr-x 2 forel geograd 512 Apr 15 11:08 data/
The left column shows that file nmoall.sh now is executable. The “*” on the file name
(right column) also indicates that the file is executable.

2.3.6 Shell Internal Variables

Inside a shell script, variables are defined with an equal (=) sign and referenced (used
in the script) with the $. The line
offset=2000
sets the variable offset to 2000. The value of the variable is later used as $offset.

2-3
 Unix Commands and Concepts

2.3.7 Passing Values to the Shell

Values can be passed to the shell when the script is called. Each argument listed after
the name of the script when it is called can be referenced in the script with the
expressions $1 (first argument), $2 (second argument), etc. The command
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

migrate1.sh 1800 4500

starts script migrate1.sh and assigns the value 1800 to variable $1 and the value 4500 to
variable $2. If migrate1.sh is a migration script, 1800 might mean a velocity of 1800 m/s,
the velocity at the top of the time section; and 4500 might mean 4500 m/s, the velocity at
the bottom of the time section. By passing values to the shell when you start the script,
you can re-run the script any number of times with different values, without rewriting the
script, until you are satisfied with the output.

2.3.8 Continuation Mark: \

Sometimes we place instructions on more than one line for the sake of visual clarity.
In the following instruction, the title of the output plot is placed on the second line. When
the continuation mark (backslash) is used, no other character or space can be to its right.
suximage < shot$1.su perc=95 \
title=”SP # $1”

2.3.9 Integer Evaluation

Integers are numbers without a decimal part. That is, 2, 0, and -1 are integers. Simple
integer expressions can be evaluated within the shell by enclosing the expression in
single back quotes (not the apostrophe). Notice that the multiply sign must use the
backslash.
var=`expr var1 [+][-][\*][/][%] var2`
For example, increment a loop counter with:
k=`expr $k + 1`
There are no spaces around the equal sign (=). However, spaces are acceptable within the
quotes.

2.3.10 Floating Point Evaluation

Floating point numbers (sometimes called “floats”) are numbers with a decimal part.
That is, 2., 0.01, and -1.5 are floating point numbers. Floating-point arithmetic is possible
using the Unix calculator bc. For example:
src=`bc -l << -END
$i * 0.05
END`
Here, src is the variable. The value of src is computed in the second line. Most of the
first line and the third line are bc syntax that must be used exactly as shown.
Note: Strictly speaking, the characters on the third line must be the only characters on
the line and must start in column 1. However, modern operating systems are not as strict
as older ones. Scripts in this Primer ignore this outdated, strict rule. However, beware!
More sophisticated script languages like awk or Perl could also be used to create
scripts with floating point arithmetic, but this would make the use of SU unnecessarily
complicated in the context of this Primer.

2-4
 Unix Commands and Concepts

2.3.11 Debug Option

Used as the first command of a script after initializing the shell, the debugging
command prints all executed lines to the screen as they execute. The screen output can be
obscure, but it can also be helpful. The command is:
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

set -x

2.3.12 Case Option

Case involves pattern matching. If pattern1 matches input variable var, commands in
list1 are executed. If, instead, pattern2 matches var, commands in list2 are executed. Etc.
case [var] in
[pattern1]) [list1];;
[pattern2]) [list2];;
... ... ;;
esac
An important part of case is the right parenthesis after the key words pattern1,
pattern2, etc. Also, notice the double semicolons at the end of each list.

2.3.13 Shell Loops and If Blocks

The programming language of the Bourne shell allows making for loops:
for [var]
do
[list]
done
do-while loops:
while [condition]
do
[list]
done
and if-blocks:
if [condition]
then
[list1]
else
[list2]
fi

2.3.14 Exit the Shell

At the end of a script the command exit can be used. If exit is used, the shell
terminates at this point, even if there are commands after exit. The command is:
exit

2.4 Shell Script Example: myplot.sh

This example shows the syntax of a “case” script and demonstrates SU plot options.
We placed line numbers on the left for discussion; they are not part of the script.
1 #! /bin/sh
2 # File: myplot.sh
3 # Plot test of "case"
4

2-5
 Unix Commands and Concepts

5 # Set messages on
6 ##set -x
7
8 # Define variable
9 signaltonoise=10
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

10
11 # Create seismic data. By default, suplane generates
12 # 32 traces with signals from three reflectors.
13 suplane | suaddnoise sn=$signaltonoise > myplot.su
14
15 # Send file myplot.su to user-selected image program
16 # or make Postscript file.
17 case $1 in
18
19 wiggle)
20 suxwigb < myplot.su title="Wiggle plot"
21 ;;
22
23 image)
24 suximage < myplot.su title="Bitmap plot"
25 ;;
26
27 pswiggle)
28 supswigp < myplot.su > myplot1.eps title="Postscript Wiggle"
29 echo " "
30 echo " Wiggle file myplot1.eps has been created."
31 echo " "
32 ;;
33
34 psimage)
35 supsimage < myplot.su > myplot2.eps title="Postscript Bitmap"
36 echo " "
37 echo " Bitmap file myplot2.eps has been created."
38 echo " "
39 ;;
40
41 *)
42 echo " "
43 echo " Use: myplot.sh [wiggle, image, pswiggle, psimage]"
44 echo " "
45 ;;
46
47 esac
48
49 # Exit politely from shell
50 exit
51

(For curious readers, we placed numbers in the left by entering:

cat -n myplot.sh > myplot.txt
Option –n of cat made the numbers. We redirected the cat output to a text file that we
later inserted into this document.)
Line 1 starts the shell. Line 2 is the first comment. Line 9 defines the signal-to-noise
ratio of the output seismic data. The defined variable signaltonoise is referenced in line
13 as $signaltonoise. (Refer to Section 2.3.6.)
Line 13: program suplane creates seismic data, the pipe (|) sends the output of
suplane to program suaddnoise, program suaddnoise adds the previously defined
amount of white (random) noise, the modified seismic data is redirected (>) to output as

2-6
 Unix Commands and Concepts

file myplot.su. By the end of line 13, you have a new file of seismic data on your
computer. File myplot.su is the input data to the “case” logic that follows.
Lines 17-47 are the “case” logic.
Line 50 exits the shell.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

To run this script, it first must be created in an editor, saved under any name and
made executable. The script is then called with the name of the file, followed by the case
selection (see Table 2.2). For example, if the script file was saved under the name
myplot.sh, and then made executable with the command
chmod +x myplot.sh
the command
myplot.sh wiggle
runs the script and displays wiggle traces on the screen. In fact, there are four ways to run
this script:
Table 2.2: Possible cases and output of myplot.sh
Command Output
myplot.sh wiggle wiggle trace screen image
myplot.sh image bitmap screen image
myplot.sh pswiggle wiggle trace Postscript file – myplot1.eps
myplot.sh psimage bitmap Postscript file – myplot2.eps
The first two cases display seismic data directly on the screen. The latter two cases
save the created image as a Postscript file. When you run the script selecting one of the
Postscript cases, you must use one of the commands below to see the contents of the
Postscript files (if you have the Ghostscript program):
ghostview –bg white myplot1.eps &
ghostview -bg white myplot2.eps &
Your output should look like one of the four images in Figures 2.1 and 2.2 (below).
Option “-bg white” makes the ghostview background white. On many systems, the
default background is grey.
The echo command writes to the screen (unless the output of echo is redirected).
Lines 29 and 31 put blank lines above and below the output of line 30 to make line 30
output easier to read. The same is true for lines 36, 38, 37 and lines 42, 44, 43.
Last, notice line 41. This case option is used if you do not select or if you incorrectly
type one of the other cases. Case option “*)” is placed after all other cases and is always
selected if none of the previous cases are selected. This option is used here to print
information to the screen to remind you of the acceptable cases.

2-7
 Unix Commands and Concepts
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 2.1: Left: case = wiggle. Right: case = image.

Figure 2.2: Left: case = pswiggle. Right: case = psimage.

2-8
 Trace Headers and Windowing Data

3. Trace Headers and Windowing Data

3.1 What are Trace Headers?
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

When seismic traces are in SEG-Y format, the SU trace format, and
many other formats, the beginning of every trace, the trace header, has
information about the trace. You can think of these as slots of
information above the data part of the trace. The data part is the time-
amplitude series that we see in a seismic display.
Trace header information might include the trace number and the
offset of the trace (for shot or CMP gathers). In an SU seismic data set,
the number of trace header slots is the same to ensure that every trace in
a data set has the same length (in terms of bytes of storage).
SU doesn't call them headers; it calls them keys. The following table
lists some SU keys.
Table 3.1: Some SU trace “headers” or trace “keys”
Key Definition
dt sample interval in microseconds
ns number of samples in this trace
ntr number of traces
offset offset
tracf trace number within field record (gather)
tracl trace sequence number within line
tracr trace sequence number within reel (entire data set)
delrt delay recording time in milliseconds

3.2 First Look at Trace Headers (keys): surange and sugethw

You can use program surange to learn which keys are in a data set and the range of
their values (the largest and smallest values). Now that you have run script myplot.sh
(Section 2.4), use program surange to examine the trace headers of the .seismic file.
Enter:
surange < myplot.su
The output (screen display) is:
32 traces:
tracl=(1,32) tracr=(1,32) offset=400 ns=64 dt=4000
This synthetic data set has 32 traces. Only five of its keys (trace headers) have non-zero
values. The minimum value of tracl and tracr is 1; the maximum value of tracl and tracr
is 32. This is not surprising since there are 32 traces in the data set. We can suppose that
the traces are numbered in sequence, 1 to 32. Surprisingly, all traces have the same offset:
400.

3-1
 Trace Headers and Windowing Data

The sample interval is 4 ms (dt = 4000 microseconds) and every trace has 64 samples
(ns = 64). So, the traces must be 256 ms long. We can see this is true in Figure 1.4.
Let’s use another trace header analysis program, sugethw, a program that gets header
words and writes them to the screen. The command below sends seismic data file
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

myplot.su to program sugethw. Keys tracl, tracr, offset, ns, and dt are read from the file.
However, instead of writing the key values to the screen, the key values are directed to a
new file, test.txt. We can see the contents of test.txt by using the cat command.
sugethw < myplot.su key=tracl,tracr,offset,ns,dt > test.txt
cat test.txt
Below are the contents of file test.txt. (Blank lines were removed from the file.)
tracl=1 tracr=1 offset=400 ns=64 dt=4000
tracl=2 tracr=2 offset=400 ns=64 dt=4000
tracl=3 tracr=3 offset=400 ns=64 dt=4000
tracl=4 tracr=4 offset=400 ns=64 dt=4000
tracl=5 tracr=5 offset=400 ns=64 dt=4000
tracl=6 tracr=6 offset=400 ns=64 dt=4000
tracl=7 tracr=7 offset=400 ns=64 dt=4000
tracl=8 tracr=8 offset=400 ns=64 dt=4000
tracl=9 tracr=9 offset=400 ns=64 dt=4000
tracl=10 tracr=10 offset=400 ns=64 dt=4000
tracl=11 tracr=11 offset=400 ns=64 dt=4000
tracl=12 tracr=12 offset=400 ns=64 dt=4000
tracl=13 tracr=13 offset=400 ns=64 dt=4000
tracl=14 tracr=14 offset=400 ns=64 dt=4000
tracl=15 tracr=15 offset=400 ns=64 dt=4000
tracl=16 tracr=16 offset=400 ns=64 dt=4000
tracl=17 tracr=17 offset=400 ns=64 dt=4000
tracl=18 tracr=18 offset=400 ns=64 dt=4000
tracl=19 tracr=19 offset=400 ns=64 dt=4000
tracl=20 tracr=20 offset=400 ns=64 dt=4000
tracl=21 tracr=21 offset=400 ns=64 dt=4000
tracl=22 tracr=22 offset=400 ns=64 dt=4000
tracl=23 tracr=23 offset=400 ns=64 dt=4000
tracl=24 tracr=24 offset=400 ns=64 dt=4000
tracl=25 tracr=25 offset=400 ns=64 dt=4000
tracl=26 tracr=26 offset=400 ns=64 dt=4000
tracl=27 tracr=27 offset=400 ns=64 dt=4000
tracl=28 tracr=28 offset=400 ns=64 dt=4000
tracl=29 tracr=29 offset=400 ns=64 dt=4000
tracl=30 tracr=30 offset=400 ns=64 dt=4000
tracl=31 tracr=31 offset=400 ns=64 dt=4000
tracl=32 tracr=32 offset=400 ns=64 dt=4000
As we suspected, the traces are numbered successively. As we knew, all traces have
offset value 400.

3.3 Real Seismic Shot Gathers

Thanks to the Center for Wave Phenomena (CWP) at the Colorado School of Mines
(CSM), we can download any of forty shot gathers collected from around the world.
These are described in Yilmaz (1989) in Table 1-8 and pages 26-41, and Yilmaz (2001)
in Table 1-13 and pages 67-80. You can download these files from:
ftp://ftp.cwp.mines.edu/pub/data/oz.original/

3-2
 Trace Headers and Windowing Data

Be sure to read the README file on this site:

ftp://ftp.cwp.mines.edu/pub/data/oz.original/README
The following sections show how to download shot gather 3 from the CWP web site
and test the gather’s “endianness.” Endianness refers to whether the downloaded binary
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

file is big-endian (high byte) or little-endian (low byte). You don’t have to understand
endianness to understand the following sections. If you want to understand it, put
“endianness” in a web search engine.

3.3.1 Get the Data

Go to the web site cited earlier and right-mouse click on ozdata.3. A pull-down menu
appears. If you are using Netscape or Mozilla, click on “Save Link Target As…” If you
are using Internet Explorer, click on “Save Target As…” If you are using a different
browser, select a similar item from the right mouse pull-down menu. Save the file to your
seismic data directory.

3.3.2 Test the Data

3.3.2.1 Big-Endian (High Byte) Machine: Solaris

At Michigan Tech, I (Forel) use a Solaris computer. According to the README file
cited earlier, Solaris is a big-endian machine, the same endianness as the data on the
CWP web site. Below, I test the file by using surange. (We first saw surange in Section
3.2). My computer prompt is “/home/forel/suscripts/data/>.”
1 /home/forel/suscripts/data/> surange < ozdata.3
2 24 traces:
3 tracl=(1,24) tracr=(1,24) fldr=10003 tracf=(1,24) cdp=(3,26)
4 cdpt=1 trid=1 nvs=1 nhs=1 duse=1
5 scalel=1 scalco=1 counit=1 delrt=4 muts=4
6 ns=1550 dt=4000
7 /home/forel/suscripts/data/> mv ozdata.3 oz03.su
8 /home/forel/suscripts/data/>
In this test:
• In line 1, I input the data to surange.
• Lines 2-6 are the output from surange. Program surange did not have any
problem reading the file.
• In line 7, I rename the file so it has our naming convention.

3.3.2.2 Little-Endian (Low Byte) Machine: PC

At home, I (Forel) use a Linux computer. According to the README file cited
earlier, my PC is a little-endian machine, NOT the same endianness as the data on the
CWP web site. Again, I test the file by using surange. My computer prompt is
“/home/david/suscripts/data/$.”
1 /home/david/suscripts/data/$ surange < ozdata.3
2
3 surange: fgettr.c: on trace #2 number of samples in header (212)
4 differs from number for first trace (32520)
5 /home/david/suscripts/data/$ suswapbytes < ozdata.3 | surange
6 24 traces:
7 tracl=(1,24) tracr=(1,24) fldr=10003 tracf=(1,24) cdp=(3,26)

3-3
 Trace Headers and Windowing Data

8 cdpt=1 trid=1 nvs=1 nhs=1 duse=1

9 scalel=1 scalco=1 counit=1 delrt=4 muts=4
10 ns=1550 dt=4000
11 /home/david/suscripts/data/$ suswapbytes < ozdata.3 > oz03.su
12 /home/david/suscripts/data/$ surange < oz03.su
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

13 24 traces:
14 tracl=(1,24) tracr=(1,24) fldr=10003 tracf=(1,24) cdp=(3,26)
15 cdpt=1 trid=1 nvs=1 nhs=1 duse=1
16 scalel=1 scalco=1 counit=1 delrt=4 muts=4
17 ns=1550 dt=4000
18 /home/david/suscripts/data/$
In this test:
• In line 1, I input the file to surange.
• Lines 3-4 show that surange cannot read the file.
• In line 5, because I suspect the problem is the endianness of the file, I use
suswapbytes to change the file’s endianness, then I pipe the altered file to
surange.
• Lines 6-10 show that surange can read the altered file.
• In line 11, I use suswapbytes to change the endianness of the input file, then I
redirect the output to a file with our naming convention.
• In line 12, I test the new file with surange.
• Lines 13-17 show that the new file can be read by surange.

3-4
 Trace Headers and Windowing Data

3.4 Program suxwigb and Option perc

Download file ozdata.3 from the CWP ftp site. Use our file naming convention –
change the file name to oz03.su by using the Unix mv command:
mv ozdata.3 oz03.su
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Look at the data by using the suxwigb program:

suxwigb < oz03.su perc=95
The perc option is useful for examining real data. (Synthetic data are usually evenly
balanced.) The perc option stands for “percentile for determining clip.” Its practical use is
to gain all traces by approximately the same amount. Try various value of perc, perhaps
from 80 to 99. If perc is not specified, suxwigb uses the default value perc=100.

Figure 3.1: Data set oz03.su displayed with perc=95.

3.5 Second Look at Trace Headers (keys): sukeyword

Below is the surange output of file oz03.su.
surange < oz03.su
24 traces:
tracl=(1,24) tracr=(1,24) fldr=10003 tracf=(1,24) cdp=(3,26)
cdpt=1 trid=1 nvs=1 nhs=1 duse=1
scalel=1 scalco=1 counit=1 delrt=4 muts=4
ns=1550 dt=4000
Again, program surange is a rapid way to see the smallest/largest values of the non-zero
value keys in a data set. (If all the values of a key are zero, surange does not print the
key.) These traces have different cdp values, and there are no offset values. (You can use
sugethw to confirm this.)
Thought points:

3-5
 Trace Headers and Windowing Data

• We do not expect dt, the sample interval key, to have more than one value.
• We expect tracf, the trace number key, to exhibit many values.
For a complete list of SU keys, use program sukeyword:
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

sukeyword -o
This program prints a document, one page at a time to the screen. Press the space bar
repeatedly to scroll to the next pages of the listing. The “–o” flag tells the program that
prints the contents of file segy.h to start the screen print with the first key definition.
Alternatively, you can send the listing to a file, like this:
sukeyword -o > sukeys.txt
Now you can either open file sukeys.txt (I made up the name.) in a text editor for
convenient reading or you can use more to scroll through the file:
more sukeys.txt
If you know the name of a key and want its definition, use the following command:
sukeyword [ ]
For example, when you type:
sukeyword dt
the screen print of file segy.h opens and quickly scrolls to put the definition of dt just a
few lines below the top of the screen. Try it!

3.6 Third Look at Trace Headers (keys): suxedit

Program suxedit is another way to learn which keys are in a data set. Program
suxedit is interactive. When you enter:
suxedit < oz03.su
the keys and their values are immediately listed for the first trace that is read. (Note:
Sometimes traces are numbered in reverse order. When that happens, the first trace read
is the one with the largest trace number.) You can then go forward or backward through
the data set to see key values of other traces. When you are in suxedit, enter "?" to get a
list of options.

3.7 Trace Header Manipulation: Constant and Regularly Varying Values

Let’s go beyond trace header viewing. Let’s modify some trace headers in oz03.su.
We saw in Section 3.5 that the traces in oz03.su have different cdp values, and all
offset values are zero. To use some of our processes (like velocity analysis), we want the
gather to appear to be a CMP (CDP). So, all the traces in the gather must have the same
value in key cdp.
We can solve the CDP problem with a one-line command that uses program sushw:
set one or more header words. This command replaces all cdp values with a single value
and creates a new seismic file oz03h1.su.
sushw < oz03.su key=cdp a=3 > oz03h1.su
See the selfdoc for sushw.
Now let's consider how to put offset values into the trace headers. First, we want to
learn what the offset values should be. Yilmaz describes this shot record in Table 1-8 of
Seismic Data Processing (1987) and in Table 1-13 of Seismic Data Analysis (2001):

3-6
 Trace Headers and Windowing Data

Table 3.2: Shot gather 3

Number of Sampling Number Trace Inner
Samples Interval of Interval Offset
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

per Trace (ms) Traces (ft) (ft)

1500 4 24 340 340
We see from Figure 3.1 that offset increases with trace number (header tracl, tracr, or
tracf.)
The following one-line command generates the appropriate offset values and creates a
new seismic file oz03h2.su.
sushw < oz03h1.su key=offset a=340 b=340 j=24 > oz03h2.su
Let's use surange to examine the trace headers:
surange < oz03h2.su
The surange output is:
24 traces:
tracl=(1,24) tracr=(1,24) fldr=10003 tracf=(1,24) cdp=3
cdpt=1 trid=1 nvs=1 nhs=1 duse=1
offset=(340,8160) scalel=1 scalco=1 counit=1 delrt=4
muts=4 ns=1550 dt=4000
Let's look at some of the trace header information:
1. tracl, tracr, tracf: Traces are numbered 1 to 24.
2. cdp: Trace cdp numbers are the single value "3.”
3. offset: The offsets range from 340 to 8160 (24 traces x 340 ft = 8160 ft).
4. dt: The sample interval is 4 ms.
5. delrt: The time delay to the first sample is 4 ms (one sample).
Note: You can use sugethw to confirm these key values.

3-7
 Trace Headers and Windowing Data

3.8 Trace Header Manipulation 2: Reverse Numbering

Download file ozdata.30 from the CWP ftp site. Use our file naming convention –
change the file name to oz30.su by using the Unix mv command:
mv ozdata.30 oz30.su
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Look at the data by using the suxwigb program:

suxwigb < oz30.su perc=85

Figure 3.2: Data set oz30.su displayed with perc=85.

Below is the surange output of this file.
surange < oz30.su
96 traces:
tracl=(1,96) tracr=(1,96) fldr=10030 tracf=(1,96) cdp=(30,125)
cdpt=1 trid=1 nvs=1 nhs=1 duse=1
scalel=1 scalco=1 counit=1 delrt=4 muts=4
ns=2175 dt=4000
Again, traces have different cdp values, and there are no offset values. And, to use
some of our processes we want the gather to appear to be a CMP (CDP). So, all the traces
in the gather must have the same value in key cdp.
We already know how to replace all cdp values with a single value, but we show it
again for fun.
sushw < oz30.su key=cdp a=30 > oz30h1.su
Before we put offset values into the trace headers, we want to learn what the offset
values should be. Again, we refer to Table 1-8 of Seismic Data Processing (Yilmaz,
1987) or Table 1-13 of Seismic Data Analysis (Yilmaz, 2001):
Table 3.3: Shot gather 30

3-8
 Trace Headers and Windowing Data

Number of Sampling Number Trace Inner

Samples Interval of Interval Offset
per Trace (ms) Traces (m) (m)
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

2125 4 96 25 230
We see from Figure 3.2 that offset decreases with increasing trace number. Before we
use sushw, we have to calculate the offset of the first trace, the farthest offset trace:
farthest offset = 230 + ( 95 * 25 ) = 2605
The following one-line command generates the appropriate offset values and creates a
new seismic file oz30h2.su.
sushw < oz30h1.su key=offset a=2605 b=-25 j=96 > oz30h2.su
Notice that a, the first trace, is assigned the value of the farthest offset and we use the
negative trace interval in b to count down from the farthest offset.
Let's use surange to examine the trace headers:
surange < oz30h2.su
The surange output is:
96 traces:
tracl=(1,96) tracr=(1,96) fldr=10030 tracf=(1,96) cdp=30
cdpt=1 trid=1 nvs=1 nhs=1 duse=1
offset=(230,2605) scalel=1 scalco=1 counit=1 delrt=4
muts=4 ns=2175 dt=4000
Let's look at some of the trace header information:
1. tracl, tracr, tracf: Traces are numbered 1 to 96.
2. cdp: Trace cdp numbers are the single value "30".
3. offset: The offsets range from 230 to 2605.
4. dt: The sample interval is 4 ms.
5. delrt: The time delay to the first sample is 4 ms (one sample).
Note: You can use sugethw to confirm these key values.
There are richer (more complex) ways to use sushw than we have shown. We leave
that exploration to you.

3-9
 Trace Headers and Windowing Data

3.9 Windowing Data

The steps below give the details of how the data/figures below were obtained.
Step 1 We "select" or "window" a single trace, 60, from original file oz30.su.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

suwind < oz30.su key=tracl min=60 max=60 > oz3060.su

Step 2 Now trace 60 is the only data in new file oz3060.su (the left part of Figure
3.3). Display the file as:
suxwigb < oz3060.su &
Step 3 From the file with single trace 60, we window the time interval of 1 to 3
seconds.
suwind < oz3060.su tmin=1 tmax=3 > oz3060t.su
Step 4 We create the right wiggle trace display to show oz3060t.su.
suxwigb < oz3060t.su &
Step 5 We list the files to see relative sizes.
ls -lF oz30*.su
858240 Dec 7 16:58 oz30.su
8940 Dec 7 17:09 oz3060.su
2244 Dec 7 17:13 oz3060t.su
Step 6 We use surange to examine trace headers of the single trace.
surange < oz3060.su
1 traces:
tracl=60 tracr=60 fldr=10030 tracf=60 cdp=89
cdpt=1 trid=1 nvs=1 nhs=1 duse=1
scalel=1 scalco=1 counit=1 delrt=4 muts=4
ns=2175 dt=4000
Step 7 We use surange to examine trace headers of the time-windowed trace..
surange < oz3060t.su
1 traces:
tracl=60 tracr=60 fldr=10030 tracf=60 cdp=89
cdpt=1 trid=1 nvs=1 nhs=1 duse=1
scalel=1 scalco=1 counit=1 delrt=1000 muts=4
ns=501 dt=4000
Step 8 Examine both surange runs. Notice that header delrt (delay recording
time in milliseconds) retains the old time annotation. That is, although
oz3060t.su is a new file, header delrt tells the display program how to
annotate the windowed segment.
Thought 1: In Step 1, if the input file contains many gathers, this step would create a
file of every trace 60 from all the gathers.
Thought 2: In Step 3, if the input gather has many traces, this step would create a file
of every trace time-limited from 1 to 3 seconds.

3-10
 Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

3-11
Trace Headers and Windowing Data

Figure 3.3: Left: Trace 60 from oz30.su. Right: Trace 60 windowed from 1 to 3 seconds.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

This page has been intentionally left blank

 Three Simple Models

4. Three Simple Models

4.1 Introduction
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

This chapter shows you how to create 2-D geologic models, create synthetic shot
gathers from the models, and examine the shot gathers for quality control (QC the
gathers). Chapter 5 shows you how to use the models to “acquire” (create synthetic)
seismic data sets.
Models developed in this chapter have layers that are homogeneous and isotropic.
Each layer has a single acoustic (P-wave) velocity.
This chapter and the next are computer-intensive in two ways: (1) the scripts are
complex and (2) it will probably take your computer two hours to most of a day to
generate the shot gathers. If you want to skip this complexity, skim these two chapters to
familiarize yourself with the geologic models, then simply use the synthetic data set
generated from Model 4 (Chapter 6) that accompanies this Primer. Seismic data
generated from Model 4 are processed in Chapters 7, 8, and 9.

4.2 Model 1: Earth Model – Five Layers

Our first model consists of five homogenous, isotropic layers. In the x-direction, the
model goes from zero to six kilometers. In the z-direction, the model goes from zero to
two kilometers. In Figure 4.1, as with all the model images in this Primer, the model is
drawn 1:1 (the horizontal units have the same length as the vertical units; the model is not
stretched or squeezed). It is important to see a model 1:1 to assess the complexity of
source-receiver raypaths.

Figure 4.1: Model 1: Five homogeneous, isotropic layers.

Let’s examine the script that created Model 1 and Figure 4.1. The numbers on the left
are added for discussion; they are not part of the script.
1 #! /bin/sh
2 # File: model1.sh
3
4 # Set messages on
5 set -x

4-1
 Three Simple Models

6
7 # Experiment Number
8 num=1
9
10 # Name output binary model file
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

11 modfile=model${num}.dat
12
13 # Name output encapsulated Postscript image file
14 psfile=model${num}.eps
15
16 # Remove previous .eps file
17 rm -f $psfile
18
19 trimodel xmin=0 xmax=6 zmin=0 zmax=2 \
20 1 xedge=0,6 \
21 zedge=0,0 \
22 sedge=0,0 \
23 2 xedge=0,2,4,6 \
24 zedge=0.30,0.50,0.20,0.30 \
25 sedge=0,0,0,0 \
26 3 xedge=0,2,4,6 \
27 zedge=0.55,0.75,0.45,0.55 \
28 sedge=0,0,0,0 \
29 4 xedge=0,2,4,6 \
30 zedge=0.65,0.85,0.55,0.65 \
31 sedge=0,0,0,0 \
32 5 xedge=0,2,4,6 \
33 zedge=1.30,1.30,1.60,1.20 \
34 sedge=0,0,0,0 \
35 6 xedge=0,6 \
36 zedge=2,2 \
37 sedge=0,0 \
38 kedge=1,2,3,4,5,6 \
39 sfill=0.1,0.1,0,0,0.44,0,0 \
40 sfill=0.1,0.4,0,0,0.40,0,0 \
41 sfill=0.1,0.6,0,0,0.35,0,0 \
42 sfill=0.1,1.0,0,0,0.30,0,0 \
43 sfill=0.1,1.5,0,0,0.25,0,0 > $modfile
44 ## x,z
45
46 # Create a Postscript file of the model
47 # Set gtri=1.0 to see sloth triangle edges
48 spsplot < $modfile > $psfile \
49 gedge=0.5 gtri=2.0 gmin=0 gmax=1 \
50 title="Earth Model - 5 layers [M${num}]" \
51 labelz="Depth (km)" labelx="Distance (km)" \
52 dxnum=1.0 dznum=0.5 wbox=6 hbox=2
53
54 # Exit politely from shell
55 exit
56
This script can be divided into sets:
• System: Line 1 invokes the shell, line 5 turns on messages, and line 55 exits the
shell.
• Variables:
o Line 8 lets us vary the number of each run as we perfect the script. This
number becomes part of output file names. Examples: 1, 101, 1a.
o Line 11 assigns a name to the output binary model file that is used on line 43.
The output binary model file is the prime reason for this script. Line 11 is

4-2
 Three Simple Models

an example of using curly brackets {} to concatenate text. Line 11 does not

need this level of clarity, but this is a good place to introduce the idea. The
curly brackets are necessary when we put text after a variable:
model${num}binary.dat
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

It is also good practice when using two variables next to each other:
model${variable1}${variable2}.dat
o Line 14 assigns a name to the output .eps image file that is used on line 48.
• Bookkeeping: Line 17 removes a previous image file. This line is optional. On
some systems, the program crashes if a .eps file with the same name already
exists. Usually files are overwritten!
• Program trimodel: Lines 19-43 create the model. Notice that line 44 is a
comment. We use line 44 to remind us that the first two entries of line sfill are x-z
values. The very important sfill parameter is discussed below.
• Program spsplot: Lines 48-52 create the .eps image file.
Let’s examine program trimodel. Program trimodel fills the model with triangles of
(1/velocity)2. While (1/velocity) is called “slowness,” (1/velocity)2 is called “sloth.” A
sloth is a slow-moving tree-dwelling mammal found in Central and South America.
Program trimodel’s use here can be divided into five parts:
1. Line 19 defines the model dimensions.
2. The six sets of xedge, zedge, and sedge triplets define layer boundaries and
velocity gradients. There is a requirement that each set of triplets have the same
number of values. In other words, the triplet for layer 1 that defines the top of the
model (a straight line) has (only needs) two values and each triplet (xedge, zedge,
sedge), therefore, contains two values. On the other hand, the triplet for layer 2
that defines a curved surface has more than two values.
20 1 xedge=0,6 \
21 zedge=0,0 \
22 sedge=0,0 \

23 2 xedge=0,2,4,6 \
24 zedge=0.30,0.50,0.20,0.30 \
25 sedge=0,0,0,0 \
A triplet consists of an xedge, zedge pair that is an interface control point and an
sedge value that is the velocity gradient at the control point.
Line sedge has only zeros because all layers are isotropic and homogeneous. Line
sedge would have non-zero values if we want a layer to have velocity gradients.
3. In this model that has six boundaries, there are five simple layers. Therefore, there
are five sfill lines, lines 39-43. These lines are written for isotropic, homogenous
layers, which is why most of the values are zero.
Table 4.1: Line sfill variables
x z x0 z0 s00 ds/dx ds/dz
Each x-z pair of sfill is a point in a layer. Each sfill line describes the sloth value
that fills the layer.

4-3
 Three Simple Models

Table 4.2: Values of sfill in Model 1

x (m) z (m) s00 (sloth) velocity (1/¥sloth) (m/s)
0.1 0.1 0.44 1508
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

0.1 0.4 0.40 1581

0.1 0.6 0.35 1690
0.1 1.0 0.30 1826
0.1 1.5 0.25 2000
These P-wave velocity values are not very realistic (except the first), but they
make a nice grey-scale image in the .eps file.
4. Line 38, line kedge, is necessary for any acquisition program that uses this model.
An interface that is not listed here will not be seen by a later acquisition
script.
5. Line 43 has the variable name of the output binary model file.
Program spsplot plots a triangulated sloth function as an encapsulated Postscript file.
So, we see that spsplot is designed to accompany trimodel! If you want to see the
triangles that are used to build the model, change gtri=2.0 to gtri=1.0.
On line 52, wbox=6 and hbox=2 ensure that the .eps image has the same dimensions
as the actual model (line 19). It is important for us to specify the same width and height
in different programs within the same script so images created by different programs
exactly overlap. You will see this in scripts in the second half of this chapter.
Also on line 52, dxnum and dznum specify the annotation increments in the respective
directions. These parameters are important in later ray trace and wavefront scripts.
Below are 60-trace split-spread shot gathers acquired from source positions x=2.0,
x=2.6, x=3.3, and x=3.95. Acquisition of these data is explained in Chapter 5.

Figure 4.2: Left to right: Shot 1, Shot 13, Shot 27, Shot 40 from Model 1.
Remember, you can use the self-documentation to learn more about any SU program
by entering just the program name on a line (Section 1.11).

4-4
 Three Simple Models

4.3 Model 2: Earth Model – High Velocity Intrusion

Our second model consists of three homogenous, isotropic layers with a high velocity
intrusion. In the x-direction, the model goes from zero to six kilometers. In the z-
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

direction, the model goes from zero to two kilometers.

Figure 4.3: Model 2: Three homogeneous layers with a high-velocity intrusion.

As you can see in Figure 4.3, this model has four layers. But the script for Model 2
(below), like the script for Model 1, has six interfaces. The important difference here is
that interface 3 and interface 4 share the same first x-z point. Because interfaces 3 and 4
“pinch” closed at x=3 and because they both extend to the right edge of the model, the
“sloth” fill of layer 2 flows around interfaces 3 and 4. Therefore, only four sfill lines are
needed.
1 #! /bin/sh
2 # File: model2.sh
3
4 # Set messages on
5 set -x
6
7 # Experiment Number
8 num=2
9
10 # Name output binary model file
11 modfile=model${num}.dat
12
13 # Name output encapsulated Postscript image file
14 psfile=model${num}.eps
15
16 # Remove previous .eps file
17 rm -f $psfile
18
19 trimodel xmin=0 xmax=6 zmin=0 zmax=2 \
20 1 xedge=0,6 \
21 zedge=0,0 \
22 sedge=0,0 \
23 2 xedge=0,2,4,6 \
24 zedge=0.3,0.5,0.2,0.3 \
25 sedge=0,0,0,0 \
26 3 xedge=3,3.5,4,6 \

4-5
 Three Simple Models

27 zedge=0.7,0.66,0.66,0.66 \
28 sedge=0,0,0,0 \
29 4 xedge=3,3.5,4,6 \
30 zedge=0.7,0.74,0.74,0.74 \
31 sedge=0,0,0,0 \
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

32 5 xedge=0,2,4,6 \
33 zedge=1.3,1.3,1.6,1.2 \
34 sedge=0,0,0,0 \
35 6 xedge=0,6 \
36 zedge=2,2 \
37 sedge=0,0 \
38 kedge=1,2,3,4,5,6 \
39 sfill=1,0.1,0,0,0.44,0,0 \
40 sfill=1,0.7,0,0,0.40,0,0 \
41 sfill=4,0.7,0,0,0.30,0,0 \
42 sfill=1,1.5,0,0,0.20,0,0 > $modfile
43 ## x,z
44
45 # Create a Postscript file of the model
46 spsplot < $modfile > $psfile \
47 gedge=0.5 gtri=2.0 gmin=0 gmax=1 \
48 title="Earth Model - High Vel. Intrusion [M${num}]" \
49 labelz="Depth (km)" labelx="Distance (km)" \
50 dxnum=1.0 dznum=0.5 wbox=6 hbox=2
51
52 # Exit politely from shell
53 exit
54
Table 4.3: Values of sfill in Model 2
x (m) z (m) s00 (sloth) velocity (1/¥sloth) (km/s)
1 0.1 0.44 1508
1 0.7 0.40 1581
4 0.7 0.30 1826
1 1.5 0.20 2236
In this isotropic, homogenous model, sfill supplies (fills) a single sloth value for a
layer when an x-z point is specified in the appropriate layer.
Below are 60-trace split-spread shot gathers acquired from source positions x=2.0,
x=2.6, x=3.3, and 3.95. Acquisition of these data is explained in Chapter 5.

Figure 4.4: Left to right: Shot 1, Shot 13, Shot 27, Shot 40 from Model 2.

4-6
 Three Simple Models

4.4 Model 3: Earth Model – Diffractor

Our third model consists of three homogenous, isotropic layers with an imbedded
high velocity diffractor. In the x-direction, the model goes from -1 to +5 kilometers. In
the z-direction, the model goes from 0 to 2 kilometers. See line 19.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 4.5: Model 3: Three homogeneous layers with a diffracting lens.

Like the previous models, this model has six interfaces. This time, interfaces 3 and 4
pinch closed on both sides; they extend for only 0.2 km. Again, the “sloth” fill of layer 2
flows around interfaces 3 and 4. Therefore, only four sfill lines are needed.
1 #! /bin/sh
2 # File: model3.sh
3
4 # Set messages on
5 set -x
6
7 # Experiment Number
8 num=3
9
10 # Name output binary model file
11 modfile=model${num}.dat
12
13 # Name output encapsulated Postscript image file
14 psfile=model${num}.eps
15
16 # Remove previous .eps file
17 rm -f $psfile
18
19 trimodel xmin=-1 xmax=5 zmin=0 zmax=2 \
20 1 xedge=-1,5 \
21 zedge=0,0 \
22 sedge=0,0 \
23 2 xedge=-1.,1.0,3.0,5.0 \
24 zedge=0.3,0.5,0.2,0.3 \
25 sedge=0,0,0,0 \
26 3 xedge=2.00,2.10,2.20 \
27 zedge=0.70,0.66,0.70 \
28 sedge=0,0,0 \
29 4 xedge=2.00,2.10,2.20 \
30 zedge=0.70,0.74,0.70 \
31 sedge=0,0,0 \

4-7
 Three Simple Models

32 5 xedge=-1.,1.0,3.0,5.0 \
33 zedge=1.3,1.3,1.6,1.2 \
34 sedge=0,0,0,0 \
35 6 xedge=-1,5 \
36 zedge=2,2 \
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

37 sedge=0,0 \
38 kedge=1,2,3,4,5,6 \
39 sfill=1.0,0.1,0,0,0.44,0,0 \
40 sfill=1.0,0.6,0,0,0.40,0,0 \
41 sfill=2.1,0.7,0,0,0.30,0,0 \
42 sfill=1.0,1.5,0,0,0.20,0,0 > $modfile
43 ## x,z
44
45 # Create a Postscript file of the model
46 spsplot < $modfile > $psfile \
47 gedge=0.5 gtri=2.0 gmin=0 gmax=1 \
48 title="Earth Model - Diffractor [M${num}]" \
49 labelz="Depth (km)" labelx="Distance (km)" \
50 dxnum=1.0 dznum=0.5 wbox=6 hbox=2
51
52 # Exit politely from shell
53 exit
54
Table 4.4: Values of sfill in Model 3
x (m) z (m) s00 (sloth) velocity (1/¥sloth) (km/s)
1.0 0.1 0.44 1508
1.0 0.6 0.40 1581
2.1 0.7 0.30 1826
1.0 1.5 0.20 2236
The .eps image is still made 1:1 because in line 50, the box dimensions are six units
wide by two units high.
Below are 60-trace split-spread shot gathers acquired from source positions x=1.0,
x=1.6, x=2.3, and 2.95. Acquisition of these data is explained in Chapter 5.

Figure 4.6: Left to right: Shot 1, Shot 13, Shot 27, Shot 40 from Model 3.

4-8
 Three Simple Models

4.5 Introduction to Ray Tracing and Wavefronts

The following scripts trace rays and show wavefronts through the models discussed
earlier. These images can help us select parameter values for acquiring full seismic data
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

sets with the models. Mostly, the images are just fun to see.
The scripts are adapted from the examples in the SU demonstration directories
demos/Synthetic/Tri/Models and demos/Synthetic/Tri/Rays. Our scripts create .eps
images, then overlay them. The scripts are called “psmerge” because they are named for
the psmerge program that merges .eps files. They all start with a .eps model file
previously created.
In brief:
• We pick a location for a single shot,
usually at the surface (z=0.0).
• We decide the number of rays that
emanate from the source, defined by
“number of angles” (nangle).
• We decide the angle of the fan of rays
that emanate from the source, defined
by “first angle” (fangle) and “last
angle” (langle).
In the example above, nangle=12, fangle=-30, and langle=45. The larger the value of
nangle, the longer it takes to produce the image. However, a complicated model might
require many rays to ensure that some will reflect to surface receivers.
• We also decide whether the rays transmit (refract) through a layer or reflect from
it. For example:
o refseq=2,0 means the rays refract through interface 2
o refseq=5,1 means the rays reflect at interface 5
Because .eps files are merged, they should have the same height and width. Also, the
z-axis label and x-axis label of ray and wavefront images either should match the labels
of the model image or the labels should not be generated. In the scripts below, labels are
not generated for ray and wavefront images since the labels are already on the model
image.
Axes labels are specified for ray and wavefront images because the default values
might not match the values generated for the model image. When the same values are
specified for each image, they overlap!
The main program, triray, generates the ray and wavefront information. It is up to us
whether we save this information. Program psgraph is called once to map the ray
information to a .eps image; it is called a separate time to map the wavefront information
to a separate .eps image. Program psmerge is used once to merge however many .eps
images were created.

4-9
 Three Simple Models

4.6 Model 1 Ray Tracing: psmerge1a.sh

Using Model 1, a source at x=4.5, z=0.0 (line 25) generates 10 rays (nangle) from
angles -45° to -10° (fangle, langle) (line 20). In addition to the final .eps merge file, only
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

ray files are output (lines 14-16).

Figure 4.7: Rays generated through Model 1.

We used the refseq parameter to specify that the rays should refract through interfaces
2, 3, and 4, but should reflect from interface 5 (line 27). Remember that the top of the
model is interface 1 and the bottom is interface 6. The input binary model file (line 11)
carries sloth information for the model.
In the model file (Section 4.2), x-z increments are specified in line 52. To override the
default values of psgraph, the x-z increments are again specified in line 33.
Unfortunately, the variables in the two programs do not have the same names.
Table 4.5: z-x increment annotation parameters
program z-increment x-increment
trimodel dznum dxnum
psgraph d1num d2num
While hbox=2.0 and wbox=6.0 ensure that the model and ray images have the same
dimensions (line 31 in this script and line 52 in the model script), psgraph still must be
told the x and z extents of the image (line 32).
1 #! /bin/sh
2 # File: psmerge1a.sh
3
4 # Set messages on
5 set -x
6
7 # Experiment number
8 num=1
9
10 # Input files
11 modelfile=model${num}.dat
12 modelpsfile=model${num}.eps

4-10
 Three Simple Models

13
14 # Output files
15 rayendsfile=rayends${num}a.dat
16 rayfile=ray${num}a.dat
17 raypsfile=ray${num}a.eps
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

18 psmergefile=psmerge${num}a.eps
19
20 # Assign values to variables
21 nangle=10 fangle=-45 langle=-15 nxz=301
22
23 # Shoot the rays
24 triray < $modelfile > $rayendsfile rayfile=$rayfile \
25 nangle=$nangle fangle=$fangle langle=$langle \
26 xs=4.5 zs=0.0 nxz=$nxz \
27 refseq=2,0 refseq=3,0 refseq=4,0 refseq=5,1
28
29 # Plot the rays
30 psgraph < $rayfile > $raypsfile \
31 nplot=`cat outpar` n=$nxz hbox=2.0 wbox=6.0 \
32 x1beg=0.0 x1end=2.0 x2beg=0 x2end=6 \
33 d1num=0.5 d2num=1.0 style=seismic linegray=0
34
35 # Merge model + rays
36 psmerge in=$modelpsfile in=$raypsfile > $psmergefile
37
38 # Exit politely from shell
39 exit
40
The rays are black because linegray=0 in psgraph (line 32).
This script (psmerge1a.sh) and the next script (psmerge1b.sh) both use files created
during our earlier work with Model 1. The input files, lines 11 and 12 of psmerge1a.sh
and psmerge1b.sh, both use variable ${num} where num=1. The output files of
psmerge1a.sh use variable ${num}a to distinguish the output files of this script from the
output files of the next script.

4.7 Model 1 Wavefronts: psmerge1b.sh

To add wavefronts to Model 1, several wavefront files are created (lines 17-19) and
psgraph is used a second time to map the wavefronts. Wavefronts are not actually drawn
on the image. Instead, marks are placed at appropriate places on the rays. To make the
illusion of wavefronts, the number of rays is considerably increased from the previous
script. Here, nangle=70 (line 23).

4-11
 Three Simple Models
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 4.8: Rays and wavefronts generated through Model 1.

Parameter nt (line 24) specifies the number of time increments to draw on the
advancing wavefront. Program transp (line 40) is necessary to transpose the wavefile
image created by triray.
1 #! /bin/sh
2 # File: psmerge1b.sh
3
4 # Set messages on
5 set -x
6
7 # Experiment number
8 num=1
9
10 # Input files
11 modelfile=model${num}.dat
12 modelpsfile=model${num}.eps
13
14 # Output files
15 rayendsfile=rayends${num}b.dat
16 rayfile=ray${num}b.dat
17 raypsfile=ray${num}b.eps
18 wavefile=wave${num}b.dat
19 wavetrans=wavetrans${num}b.dat
20 wavepsfile=wave${num}b.eps
21 psmergefile=psmerge${num}b.eps
22
23 # Assign values to variables
24 nangle=70 fangle=-45 langle=-15 nxz=301 nt=20
25
26 # Shoot the rays
27 triray < $modelfile > $rayendsfile \
28 rayfile=$rayfile wavefile=$wavefile \
29 nangle=$nangle fangle=$fangle langle=$langle \
30 nxz=$nxz nt=$nt xs=4.5 zs=0.0 \
31 refseq=2,0 refseq=3,0 refseq=4,0 refseq=5,1
32
33 # Plot the rays
34 psgraph < $rayfile > $raypsfile \
35 nplot=`cat outpar` n=$nxz hbox=2.0 wbox=6.0 \
36 x1beg=0.0 x1end=2.0 x2beg=0 x2end=6 \
37 d1num=0.5 d2num=1.0 style=seismic linegray=1

4-12
 Three Simple Models

38
39 # Transpose the wavefile
40 transp < $wavefile > $wavetrans n1=$nt n2=$nangle nbpe=8
41
42 # Plot the wavefronts
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

43 psgraph < $wavetrans > $wavepsfile \

44 linewidth=0.0 mark=8 marksize=2 \
45 nplot=$nt n=$nangle hbox=2.0 wbox=6.0 \
46 x1beg=0.0 x1end=2.0 x2beg=0.0 x2end=6.0 \
47 d1num=0.5 d2num=1.0 style=seismic linegray=0
48
49 # Merge model + rays + wavefronts
50 psmerge in=$modelpsfile in=$raypsfile in=$wavepsfile > $psmergefile &
51
52 # Exit politely from shell
53 exit
54
Each time psgraph is used, the image dimensions must be specified the same way
(parameters hbox, wbox, x1beg, x1end, x2beg, x2end, d1num, d2num) to ensure that all
merged images (line50) overlap exactly.
Now the rays are white due to linegray=1 in line 37. The wavefronts are black due to
linegray=0 in line 47. Parameter linegray controls a gray continuum between 0 and 1.

4.8 Model 2 Ray Tracing: psmerge2a.sh

Using Model 2, a source at x=3.0, z=0.0 (line 25) generates 15 rays (nangle) from
angles 0° to 50° (fangle, langle) (line 20). In addition to the final .eps merge file, only ray
files are output (lines 14-16). We see some bending as rays pass through the intrusion.

Figure 4.9: Rays generated through Model 2.

The refseq parameter specifies that the rays should refract through interfaces 2, 3, and
4, but should reflect from interface 5 (line 26). The input binary model file (line 10)
carries sloth information for the model.
1 #! /bin/sh
2 # File: psmerge2a.sh
3
4 # Set messages on
5 set -x
6

4-13
 Three Simple Models

7 # Experiment number
8 num=2
9
10 # Input files
11 modelfile=model${num}.dat
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

12 modelpsfile=model${num}.eps
13
14 # Output files
15 rayendsfile=rayends${num}a.dat
16 rayfile=ray${num}a.dat
17 raypsfile=ray${num}a.eps
18 psmergefile=psmerge${num}a.eps
19
20 # Assign values to variables
21 nangle=15 fangle=0 langle=50 nxz=301
22
23 # Shoot the rays
24 triray < $modelfile > $rayendsfile rayfile=$rayfile \
25 nangle=$nangle fangle=$fangle langle=$langle \
26 xs=3.0 zs=0.0 nxz=$nxz \
27 refseq=2,0 refseq=3,0 refseq=4,0 refseq=5,1
28
29 # Plot the rays
30 psgraph < $rayfile > $raypsfile \
31 nplot=`cat outpar` n=$nxz hbox=2 wbox=6 \
32 x1beg=0 x1end=2 x2beg=0 x2end=6 \
33 d1num=0.5 d2num=1.0 style=seismic linegray=0
34
35 # Merge model + rays
36 psmerge in=$modelpsfile in=$raypsfile > $psmergefile &
37
38 # Exit politely from shell
39 exit
40
The rays are black because linegray=0 in psgraph (line 32).

4.9 Model 2 Wavefronts: psmerge2b.sh

To add wavefronts to Model 2, the number of rays is considerably increased from the
previous script. Here, nangle=65 (line 24).

Figure 4.10: Rays and wavefronts generated through Model 2.

4-14
 Three Simple Models

1 #! /bin/sh
2 # File: psmerge2b.sh
3
4 # Set messages on
5 set -x
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

6
7 # Experiment number
8 num=2
9
10 # Input files
11 modelfile=model${num}.dat
12 modelpsfile=model${num}.eps
13
14 # Output files
15 rayendsfile=rayends${num}b.dat
16 rayfile=ray${num}b.dat
17 raypsfile=ray${num}b.eps
18 wavefile=wave${num}b.dat
19 wavetrans=wavetrans${num}b.dat
20 wavepsfile=wave${num}b.eps
21 psmergefile=psmerge${num}b.eps
22
23 # Assign values to variables
24 nangle=65 fangle=0 langle=50 nxz=301 nt=20
25
26 # Shoot the rays
27 triray < $modelfile > $rayendsfile \
28 rayfile=$rayfile wavefile=$wavefile \
29 nangle=$nangle fangle=$fangle langle=$langle \
30 nxz=$nxz nt=$nt xs=3.0 zs=0.0 \
31 refseq=2,0 refseq=3,0 refseq=4,0 refseq=5,1
32
33 # Plot the rays
34 psgraph < $rayfile > $raypsfile \
35 nplot=`cat outpar` n=$nxz hbox=2 wbox=6 \
36 x1beg=0 x1end=2 x2beg=0 x2end=6 \
37 d1num=0.5 d2num=1.0 style=seismic linegray=0
38
39 # Transpose the wavefile
40 transp < $wavefile > $wavetrans n1=$nt n2=$nangle nbpe=8
41
42 # Plot the wavefronts
43 psgraph < $wavetrans > $wavepsfile \
44 linewidth=0.0 mark=8 marksize=2 \
45 nplot=$nt n=$nangle hbox=2.0 wbox=6.0 \
46 x1beg=0.0 x1end=2.0 x2beg=0.0 x2end=6.0 \
47 d1num=0.5 d2num=1.0 style=seismic linegray=1
48
49 # Merge model + rays + wavefronts
50 psmerge in=$modelpsfile in=$raypsfile in=$wavepsfile > $psmergefile &
51
52 # Exit politely from shell
53 exit
54
Now the rays are black due to linegray=0 in line 36. The wavefronts are white due to
linegray=1 in line 47.

4-15
 Three Simple Models

4.10 Model 3 Ray Tracing: psmerge3.sh

Using Model 3, a source at x=2.0, z=0.0 (line 26) generates 10 rays (nangle) from
angles 0° to 25° (fangle, langle) (line 21). In addition to the final .eps merge file, only ray
files are output (lines 15-17).
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 4.11: Rays generated through Model 3. Program triray parameter nxz=301.
The source was placed over the diffractor so rays shoot through and around it. In a
real earth, raypaths would be quite scattered near the diffractor edges. Here, the edges are
not treated realistically. Notice that only six of the ten rays reach the surface. In the first
use of this script, triray parameter nxz=301 (lines 21 and 26). To make all ten rays reach
the surface, nxz was increased in increments of 50. All rays reached the surface with
nxz=701 (see Figure 4.12, below). It is interesting to observe that the default value of nxz
is 101. It is probably unrealistic to set nxz so high (701), and probably a waste of
computer time; although, without examining the source code and understanding exactly
what role nxz plays, it is not easy to be certain. Nonetheless, through trial and error we
obtained a reasonable solution..

Figure 4.12: Rays generated through Model 3. Program triray parameter nxz=701.

4-16
 Three Simple Models

While hbox=2.0 and wbox=6.0 ensure that the model and ray images have the same
dimensions (line 31 in this script and line 50 in the model script), psgraph still must be
told the x and z extents of the image (line 32). Notice that x2beg=-1 and x2end=5.
1 #! /bin/sh
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

2 # File: psmerge3.sh
3
4 # Turn on messages
5 set -x
6
7 # Experiment number
8 num=3
9
10 # Input files
11 modelfile=model${num}.dat
12 modelpsfile=model${num}.eps
13
14 # Output files
15 rayendsfile=rayends${num}.dat
16 rayfile=ray${num}.dat
17 raypsfile=ray${num}.eps
18 psmergefile=psmerge${num}.eps
19
20 # Assign values to variables
21 nangle=10 fangle=0 langle=+25 nxz=701
22
23 # Shoot the rays
24 triray < $modelfile > $rayendsfile rayfile=$rayfile \
25 nangle=$nangle fangle=$fangle langle=$langle \
26 xs=2.0 zs=0.0 nxz=$nxz \
27 refseq=2,0 refseq=3,0 refseq=4,0 refseq=5,1
28
29 # Plot the rays
30 psgraph < $rayfile > $raypsfile \
31 nplot=`cat outpar` n=$nxz hbox=2 wbox=6 \
32 x1beg=0 x1end=2 x2beg=-1 x2end=5 \
33 d1num=0.5 d2num=1.0 style=seismic linegray=0
34
35 # Merge model + rays
36 psmerge in=$modelpsfile in=$raypsfile > $psmergefile &
37
38 # Exit politely from shell
39 exit
40

4-17
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

This page has been intentionally left blank

 Three Simple Models: Acquire 2-D Lines

5. Three Simple Models: Acquire 2-D Lines

5.1 Introduction
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

In the previous chapter, we developed three simple models and used them to generate
single shot gathers and images of rays and wavefronts. In this chapter, we use the same
models to acquired 2-D lines of seismic data.

5.2 Acquire 2-D Line from Model 1: acq1.sh

Script acq1.sh acquires seismic data over model model1.dat, generated by script
model1.sh (Section 4.2). The survey layout:
1. 40 shots (line 28)
2. Shots are equally spaced at 50m intervals (line 32 and line 35)
3. 60 split-spread traces are recorded from each shot location (line 42)
4. Geophone spacing is 50m (line 46 and line 49)
5. Geophone offsets range from -1475 m to +1475 m (line 52)
6. Shot locations range from 2 km to 3.95 km (line 68)
7. Geophone locations range from 0.525 km to 5.425 km (line 68)
The final data set has 2400 traces (40 shots x 60 geophones per shot). The generation
of this data set took about half an hour on a Sun UltraSPARC III with four processors.
1 #! /bin/sh
2 # acq1.sh
3 # Set messages on
4 ##set -x
5
6 # Assign values to variables
7 num=1
8 nangle=201 fangle=-65 langle=65
9 nt=751 dt=0.004
10
11 # Name input model file
12 inmodel=model$num.dat
13
14 # Name output seismic file
15 outseis=seis$num.su
16
17 #=================================================
18 # Create the seismic traces with "triseis"
19 # i-loop = 40 source positions
20 # j-loop = 60 geophone positions (split-spread)
21 # per shot position
22 # k-loop = layers 2 through 5
23 # (do not shoot layers 1 and 6)
24
25 echo " --Begin looping over triseis."
26
27 i=0
28 while [ "$i" -ne "40" ]
29 do
30
31 fs=`bc -l <<-END
32 $i * 0.05
33 END`
34 sx=`bc -l <<-END
35 $i * 50

5-1
 Three Simple Models: Acquire 2-D Lines

36 END`
37 fldr=`bc -l <<-END
38 $i + 1
39 END`
40
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

41 j=0
42 while [ "$j" -ne "60" ]
43 do
44
45 fg=`bc -l <<-END
46 $i * 0.05 + $j *0.05
47 END`
48 gx=`bc -l <<-END
49 $i * 50 + $j * 50 - 1475
50 END`
51 offset=`bc -l <<-END
52 $j * 50 - 1475
53 END`
54 tracl=`bc -l <<-END
55 $i * 60 + $j + 1
56 END`
57 tracf=`bc -l <<-END
58 $j + 1
59 END`
60
61 echo " Sx=$sx Gx=$gx fldr=$fldr Offset=$offset tracl=$tracl\
62 fs=$fs fg=$fg"
63
64 k=2
65 while [ "$k" -ne "6" ]
66 do
67
68 triseis < $inmodel xs=2,3.95 xg=0.525,5.425 zs=0,0 zg=0,0 \
69 nangle=$nangle fangle=$fangle langle=$langle \
70 kreflect=$k krecord=1 fpeak=12 lscale=0.5 \
71 ns=1 fs=$fs ng=1 fg=$fg nt=$nt dt=$dt |
72 suaddhead nt=$nt |
73 sushw key=dt,tracl,tracr,fldr,tracf,trid,offset,sx,gx \
74 a=4000,$tracl,$tracl,$fldr,$tracf,1,$offset,$sx,$gx >> temp$k
75
76 k=`expr $k + 1`
77
78 done
79 j=`expr $j + 1`
80
81 done
82 i=`expr $i + 1`
83
84 done
85
86 echo " --End looping over triseis."
87
88 #=================================================
89
90 # Sum contents of the "temp" files
91 echo " --Sum files."
92 susum temp2 temp3 > tempa
93 susum tempa temp4 > tempb
94 susum tempb temp5 > $outseis
95
96 # Remove temp files
97 echo " --Remove temp files."
98 rm -f temp*

5-2
 Three Simple Models: Acquire 2-D Lines

99
100 # Exit politely from shell script
101 echo " --Finished!"
102 exit
103
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

This script writes messages to the screen to help us track progress:

25 echo " --Begin looping over triseis."
61 echo " Sx=$sx Gx=$gx fldr=$fldr Offset=$offset tracl=$tracl\
62 fs=$fs fg=$fg"
86 echo " --End looping over triseis."
91 echo " --Sum files."
97 echo " --Remove temp files."
101 echo " --Finished!"
Lines 61 and 62 write variable values to the screen. Notice that line 4 is commented out.
We find that once a script is perfected, the messages from line 4 interfere with reading
the values printed to the screen by lines 61 and 62.
Previously (Section 4.5), we discussed variables nangle (the number of rays or angles
that emanate from the source), fangle (the first angle of the fan of rays that emanate from
the source), and langle (the last angle of the fan of rays that emanate from the source).
Here (and in the following acquisition scripts), these values are assigned on line 8.
Line 9 assigns the number of time samples (nt) and the time sample interval in
seconds (dt).
Note: Check the unit of the time sample interval in each SU program. The unit varies.
The core of the script is the three do-while loops (Section 2.3.13) over triseis,
suaddhead, and sushw. Each time these programs are used, triseis creates a seismic
trace, suaddhead creates a trace header on the trace, and sushw writes values to the trace
headers.
• The outermost i-loop goes over 40 source positions, 0 to 39 (lines 27 and 28).
• The j-loop goes over 60 geophone positions, 0 to 59 (lines 41 and 42).
• The innermost k-loop goes over the reflectors, 2 to 5 (lines 64 and 65).
Consider the innermost loop. The value of k signifies the reflector from which triseis
records a reflection (kreflect=$k, line 70) and the temp file in which the trace is stored
(temp$k, line 74.) Because k loops over reflectors 2 to 5, four temp files are created. (The
top and base of the model are ignored.) Each temp data set has (in this case) 2400 traces.
After all traces are created, lines 93 to 95 “sum” the four data sets, trace-for-trace.
The following pseudo-code of the do-while loops is presented to help us consider
how the variables are used. The program in which the variable is used is printed in bold
to the right on the assignment line. Variable k is underlined to remind us of its use as
discussed in the preceding paragraph.
Variable fs (line E) is the source location relative to the start of the source surface.
The source surface is defined in line Y as variable xs. As i increments, the source moves
across the model. Both fs and sx are the source position. Variable fs is in model units,
kilometers, and is used by triseis; variable sx is an integer number of meters and is
computed for a header value.
Variable fg (line M) is the geophone location relative to the start of the geophone
surface. The geophone surface is defined in line Y as variable gx. As j increments, the

5-3
 Three Simple Models: Acquire 2-D Lines

geophone moves across the model. Both fg and gx are the geophone position. Variable fg
is in model units, kilometers, and is used by triseis; variable gx is an integer number of
meters and is computed for a header value.
Variable fldr (line G) is a header value that identifies each shot gather. Variable tracl
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

(line P) sequentially numbers the traces of the line (TRACe number in the Line). Variable
tracf (line Q) sequentially numbers the traces of each gather (TRACe number in the File).
A i=0
B while i ne 40
C do
D
E fs = i * 0.05 triseis
F sx = i * 50 sushw
G fldr = i + 1 sushw
H
I j=0
J while j ne 60
K do
L
M fg = i * 0.05 + j *0.05 triseis
N gx = i * 50 + j * 50 - 1475 sushw
O offset = j * 50 - 1475 sushw
P tracl = i * 60 + j + 1 sushw
Q tracf = j + 1 sushw
R
S echo " $sx $gx $fldr $offset $tracl $fs $fg"
T
U k=2
V while k ne 6
W do
X
Y triseis < $inmodel xs=2,3.95 xg=0.525,5.425 zs=0,0 zg=0,0 \
Z nangle=$nangle fangle=$fangle langle=$langle \
AA kreflect=$k krecord=1 fpeak=12 lscale=0.5 \
BB ns=1 fs=$fs ng=1 fg=$fg nt=$nt dt=$dt |
CC suaddhead nt=$nt |
DD sushw key=dt,tracl,tracr,fldr,tracf,trid,offset,sx,gx \
EE a=4000,$tracl,$tracl,$fldr,$tracf,1,$offset,$sx,$gx >> temp$k
FF
GG k = k + 1
HH
II done
JJ j = j + 1
KK
LL done
MM i= i + 1
NN
OO done
Variable offset (line O) ranges from -1475 to +1475, incrementing by 50 meters. The
values of offset as it passes through zero offset are … -125, -75, -25, +25, +75, +125, etc.
This simple numbering scheme cleverly skips the geophysically impossible zero offset
acquisition position.
Let's use surange to examine the trace headers:
surange < seis1.su
The surange output is:
2400 traces:
tracl=(1,2400) tracr=(1,2400) fldr=(1,40) tracf=(1,60) trid=1
offset=(-1475,1475) sx=(0,1950) gx=(-1475,3425) ns=751 dt=4000

5-4
 Three Simple Models: Acquire 2-D Lines

5.3 Acquire 2-D Line from Model 2: acq2.sh

Script acq2.sh acquires seismic data over model model2.dat, generated by script
model2.sh (Section 4.3). The survey layout is the same as in acq1.sh; that is:
• Shot locations range from 2 km to 3.95 km (line 68)
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

• Geophone locations range from 0.525 km to 5.425 km (line 68)

For seismic data acquired by acq2.sh, we want header (key) sx to hold actual source
locations and we want header gx to hold actual receiver locations. With this in mind, the
following are lines of acq2.sh modified from acq1.sh to put actual source and receiver
locations in the headers.
34 sx=`bc -l <<-END
35 $i * 50 + 2000
36 END`

48 gx=`bc -l <<-END
49 $i * 50 + $j * 50 + 525
50 END`
Let's use surange to examine the trace headers:
surange < seis2.su
The surange output is:
2400 traces:
tracl=(1,2400) tracr=(1,2400) fldr=(1,40) tracf=(1,60) trid=1
offset=(-1475,1475) sx=(2000,3950) gx=(525,5425) ns=751 dt=4000
We see that headers sx and gx of seis2.su differ from seis1.su. All other headers are
the same as before.
The only other ways acq2.sh differs from acq1.sh are lines 2 and 7:
2 # File: acq2.sh
7 num=2
That is, the file name is internally documented and Model 2 is called instead of Model 1.

5.4 Acquire 2-D Line from Model 3: acq3.sh

Script acq3.sh acquires seismic data over model model3.dat, generated by script
model3.sh (Section 4.4). Source and geophone positions in acq3.sh are shifted one unit to
the left relative to the previous acquisition scripts. We can see this by line 68:
68 triseis < $inmodel xs=1,2.95 xg=-0.475,4.425 zs=0,0 zg=0,0 \
The survey layout is:
• Shot locations (triseis variable xs) range from 1 km to 2.95 km
• Geophone locations (triseis variable xg) range from -0.475 km to 4.425 km
For seismic data acquired by acq3.sh, we want header (key) sx to hold actual source
locations and we want header gx to hold actual receiver locations. With this in mind, the
following are lines of acq3.sh:
34 sx=`bc -l <<-END
35 $i * 50 + 1000
36 END`

48 gx=`bc -l <<-END
49 $i * 50 + $j * 50 - 475
50 END`
Let's use surange to examine the trace headers:

5-5
 Three Simple Models: Acquire 2-D Lines

surange < seis3.su

The surange output is:
2400 traces:
tracl=(1,2400) tracr=(1,2400) fldr=(1,40) tracf=(1,60) trid=1
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

offset=(-1475,1475) sx=(1000,2950) gx=(-475,4425) ns=751 dt=4000

We see that headers sx and gx of seis3.su contain actual source and receiver locations.
All other headers are the same as before.
The only other ways acq3.sh differs from previous acquisition scripts are line 2 and 7:
2 # File: acq3.sh
7 num=3
That is, the file name is internally documented and Model 3 is called.

5.5 Computer Note 7: Pause and Resume Screen Print

While a process, such as a script, is writing messages to the screen, and it is not
running in the background (see Section 1.8.1), you can pause screen printing and resume
screen printing. These actions can be done only in an active window. You can make a
window active by clicking on the window frame.
1. Stop screen printing: hold down the Control key and briefly press the S key. That
is, press Control-S.
2. Resume screen printing: hold down the Control key and briefly press the Q key.
That is, press Control-Q.

5.6 Examine Shot Gathers: showshot.sh

Now that we have seismic data (output files seis1.su, seis2.su, and seis3.su), it is time
to extract shot gathers for review. Script showshot.sh (below):
• windows one gather from a seismic data set (line 14),
• makes a wiggle display of the seismic data (lines 17-19),
• converts the seismic data to a .eps file (lines 22-24), then
• writes the .eps image to output (the end of line 24).
The script expects two values on the command line, one for the model, another for the
shot number, as shown in lines 7 and 8.
This script assumes that all seismic data are in files of the same type of name as
shown underlined in line 14 (seis1.su, seis2.su, etc.), and in the same directory as the
script.
1 #!/bin/sh
2
3 # showshot.sh: Window one "field record" from file seis#.su
4 # where # represents the model number.
5 # Outputs: wiggle image of the shot gather
6 # .eps file of the shot gather
7 # Use: showshot.sh model shot
8 # Example: showshot.sh 3 20
9
10 # Set messages on
11 set -x
12
13 # Window one "field record" to a temporary file
14 suwind < seis$1.su key=fldr min=$2 max=$2 > temp$1$2.su
15

5-6
 Three Simple Models: Acquire 2-D Lines

16 # Make wiggle plot

17 suxwigb < temp$1$2.su title="SP # $2 [$1]" key=offset \
18 label1=" Time (s)" label2="Offset" \
19 x2beg=-1500 x2end=1500 perc=99 &
20
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

21 # Create .eps image of a shot gather

22 supswigp < temp$1$2.su title="SP # $2 [$1]" key=offset \
23 label1="Time (s)" label2="Offset (m)" \
24 x2beg=-1500 x2end=1500 perc=99 > shot$1$2.eps &
25
26 # Remove temporary gather
27 rm -f temp$1$2.su
28
29 # Exit politely from shell
30 exit
31
Entering
showshot.sh 2 25
commands showshot.sh to read file seis2.su, extract shot 25, and create file shot225.eps.
That is, the output .eps file name has the number of the model and the number of the shot,
shown underlined in line 24. The title on the output images (lines 17 and 22) contain the
abbreviation SP (shot point), the number of the shot gather, and in square brackets, the
number of the seismic file (Figure 5.2).

Figure 5.2: Image of Shot Point 25 from Model 2.

This script does not output a .su file. A temporary .su file is created on line 14, is used
by suxwigb (line 17) and supswigp (line 22), and it is removed by line 27. If you want a
permanent .su file:
• Change the word “temp” (lines 14, 17, and 22) to something more useful (like
“shot”).
• Comment out line 27 or remove lines 26 and 27.
• Add a comment line after line 5 describing the new output file.

5-7
 Three Simple Models: Acquire 2-D Lines

• Remove the word “temporary” from comment line 13

Internally, the script refers to the first variable that it reads from input, model number,
as $1 and the second variable that it reads from input, shot number, is $2.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

5.7 A Graphics Program is a Terminal Program

A graphics program (suxwigb, supswigp, etc.) is a terminal program; data cannot be
piped (“|”) from a graphics program to another program. In showshot.sh, the output of
suwind has to go to two graphics programs, so the disk file is necessary. If a script uses
only one graphics program, we can pipe the output of a processing program directly to a
graphics program without making a disk file.
1. The following is acceptable:
process | graphics program
2a. The following is NOT acceptable:
process | graphics program 1 | graphics program 2
2b. Instead, do the following:
process > disk file
graphics program 1 < disk file
graphics program 2 < disk file

5.8 Computer Note 8: Use, Modify a Command from the History List
The shell has a record of your previous commands. You can use this “history” to
repeat previous commands or recall and modify previous commands.

5.8.1 Use a Command from the History List

5.8.1.1 Up Arrow Key

Press the up arrow key repeatedly to access previous commands. When you see the
command you want to use, press “Enter” or “Return” to execute the command.

5.8.1.2 History List

Enter:
history
A numbered list of previous commands scrolls on the screen. To use a previous command
from the history list, enter the exclamation mark (!) and the number of the command,
without spaces. For example:
!8
(sometimes pronounced “bang eight” or “shriek eight”) executes command 8 from the
list.

5.8.2 Modify a Command from the History List

Instead of typing a long command to make a minor modification, you can recall a
previous command and modify it using a combination of the arrow keys, the delete key,
and the I key. Suppose your previous command was (see Section 5.6):
showshot.sh 2 25

5-8
 Three Simple Models: Acquire 2-D Lines

which made a .eps file of Model 2, shot 25. Now suppose you want to make a .eps file of
Model 3, shot 25. In other words, you want to change “2” to “3.”
1. Press the up arrow key once to return to the previous command.
2. Press .the left arrow key several times, until the prompt is between the “2” and the
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

“25.”
3. Press the delete key once. The “2” is now erased.
4. Press the I key once. This tells the shell you want to “insert” on the line.
5. Press “3.” The model number is now changed.
6. Press “Enter” or “Return” to execute the command.
If you want to make other changes on the line, like changing the shot number, use the
left or right arrow keys to move the cursor to the right of the delete/insert point. Then, use
“delete” or the “I” key to make your change.
Also note:
7. Pressing the “Escape” key cancels the “insert” mode.
8. If you are not in “insert” mode, you can press “U” to undo a previous change.
When you are finished changing the line, execute your new command.

5.9 Survey Quality Control

We can run an acquisition script quickly to learn if the computed header values are
the ones we want. We do this by writing the values computed by the shell to a file and
commenting out the SU programs. Since the SU programs are not used, the script
executes quickly. A script that normally takes half an hour to run takes only four minutes
when modified as described.
The scripts in this chapter were tested with the following modifications.
At the end of line 62, the redirect with append (“>>”) is used to write the computed
values to ASCII file survey2.txt. If redirect had been used without append (a single “>”),
only one line would be in the file because each time the script loops, new values would
overwrite old values. Append puts each output line below the previous line of output.
61 echo " Sx=$sx Gx=$gx fldr=$fldr Offset=$offset tracl=$tracl\
62 fs=$fs fg=$fg" >> survey2.txt

68 ## triseis < $inmodel xs=2,3.95 xg=0.525,5.425 zs=0,0 zg=0,0 \

69 ## nangle=$nangle fangle=$fangle langle=$langle \
70 ## kreflect=$k krecord=1 fpeak=12 lscale=0.5 \
71 ## ns=1 fs=$fs ng=1 fg=$fg nt=$nt dt=$dt |
72 ## suaddhead nt=$nt |
73 ## sushw key=dt,tracl,tracr,fldr,tracf,trid,offset,sx,gx \
74 ## a=4000,$tracl,$tracl,$fldr,$tracf,1,$offset,$sx,$gx >> temp$k

91 ##echo " --Sum files."

92 ##susum temp2 temp3 > tempa
93 ##susum tempa temp4 > tempb
94 ##susum tempb temp5 > $outseis
Since no traces are created, the script would crash if it tried to execute lines 92-94; so,
we comment them out. Line 91 is commented out because we do not want to be told that
summation occurred when we know it did not.

5-9
 Three Simple Models: Acquire 2-D Lines

Notice that lines 68-74 and lines 91-94 are commented out with two “#.” We know
only one is necessary. We use two so we see where we intend comments to be temporary.
This script would have created 2400 traces. The output file contains 2400 lines and
occupies less than 2 Mbytes. Below are the first ten lines and the last ten lines of
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

survey2.txt.
Sx=2000 Gx=525 fldr=1 Offset=-1475 tracl=1 fs=0 fg=0
Sx=2000 Gx=575 fldr=1 Offset=-1425 tracl=2 fs=0 fg=.05
Sx=2000 Gx=625 fldr=1 Offset=-1375 tracl=3 fs=0 fg=.10
Sx=2000 Gx=675 fldr=1 Offset=-1325 tracl=4 fs=0 fg=.15
Sx=2000 Gx=725 fldr=1 Offset=-1275 tracl=5 fs=0 fg=.20
Sx=2000 Gx=775 fldr=1 Offset=-1225 tracl=6 fs=0 fg=.25
Sx=2000 Gx=825 fldr=1 Offset=-1175 tracl=7 fs=0 fg=.30
Sx=2000 Gx=875 fldr=1 Offset=-1125 tracl=8 fs=0 fg=.35
Sx=2000 Gx=925 fldr=1 Offset=-1075 tracl=9 fs=0 fg=.40
Sx=2000 Gx=975 fldr=1 Offset=-1025 tracl=10 fs=0 fg=.45

Sx=3950 Gx=4975 fldr=40 Offset=1025 tracl=2391 fs=1.95 fg=4.45

Sx=3950 Gx=5025 fldr=40 Offset=1075 tracl=2392 fs=1.95 fg=4.50
Sx=3950 Gx=5075 fldr=40 Offset=1125 tracl=2393 fs=1.95 fg=4.55
Sx=3950 Gx=5125 fldr=40 Offset=1175 tracl=2394 fs=1.95 fg=4.60
Sx=3950 Gx=5175 fldr=40 Offset=1225 tracl=2395 fs=1.95 fg=4.65
Sx=3950 Gx=5225 fldr=40 Offset=1275 tracl=2396 fs=1.95 fg=4.70
Sx=3950 Gx=5275 fldr=40 Offset=1325 tracl=2397 fs=1.95 fg=4.75
Sx=3950 Gx=5325 fldr=40 Offset=1375 tracl=2398 fs=1.95 fg=4.80
Sx=3950 Gx=5375 fldr=40 Offset=1425 tracl=2399 fs=1.95 fg=4.85
Sx=3950 Gx=5425 fldr=40 Offset=1475 tracl=2400 fs=1.95 fg=4.90
Comparing the output to line 68 (triseis), notice that:
• the first Sx value matches the first xs value of line 68 (2),
• the last Sx value matches the last xs value of line 68 (3.95),
• the first Gx value matches the first xg value of line 68 (0.525), and
• the last Gx value matches the last xg value of line 68 (5.425).

5-10
 Build Model 4, Acquire a Line, Display Gathers, QC

6. Model 4: Build, Acquire a Line, Display Gathers, QC

6.1 Introduction
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

In the previous two chapters, we developed three simple models and used them to
acquire 2-D lines of seismic data. In this chapter, we combine some of the attributes of
those models to create a fourth model and use the model to acquire a 2-D line of seismic
data
In the following two sections, we explain the model and acquisition scripts in detail;
you do not have to be familiar with the related scripts that are in the previous two
chapters.
The next three sections of this chapter explain how to build the model, acquire
seismic data, and view selected gathers. While those sections are important, the last two
sections about quality control (QC) are equally important. The QC sections explain how
you can acquire survey information and seismic data from selected portions of the model.
You can examine preliminary survey information and seismic data to increase your
confidence in your model and your acquisition script before spending time acquiring the
full seismic data set.

6.2 Model 4: Earth Model – Diffractor and Pinchout

As with previous models, Model 4 layers are homogeneous and isotropic. Each layer
has a single acoustic (P-wave) velocity. Let’s examine script model4.sh. The numbers on
the left are added for discussion; they are not part of the script.
1 #! /bin/sh
2 # File: model4.sh
3
4 # Set messages on
5 set -x
6
7 # Experiment Number
8 num=4
9
10 # Name output binary model file
11 modfile=model${num}.dat
12
13 # Name output encapsulated Postscript image file
14 psfile=model${num}.eps
15
16 # Remove previous .eps file
17 rm -f $psfile
18
19 trimodel xmin=-2 xmax=12.0 zmin=0 zmax=2.0 \
20 1 xedge=-2,12 \
21 zedge=0,0 \
22 sedge=0,0 \
23 2 xedge=-2.,0.00,2.0,4.0,6.0,8.00,10.0,12.0 \
24 zedge=0.3,0.32,0.3,0.6,0.2,0.25,0.25,0.25 \
25 sedge=0,0,0,0,0,0,0,0 \
26 3 xedge=-2.,0.0,2.0,4.0,6.0,7.0,7.5,8.00,8.50,9.0,9.50,10.,10.5,11.0,11.5,12. \
27 zedge=0.8,0.8,1.0,1.3,0.5,0.5,0.7,0.9,1.02,1.15,1.25,1.25,1.2,1.15,1.08,1.0 \
28 sedge=0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 \
29 4 xedge=-2.,0.0,2.0,4.0,6.0,7.00,8.0,9.0,10.0,11.0,12. \
30 zedge=1.5,1.5,1.6,1.9,1.3,1.45,1.6,1.7,1.75,1.65,1.5 \
31 sedge=0,0,0,0,0,0,0,0,0,0,0 \
32 5 xedge=1.80,2.00,2.20 \
33 zedge=0.98,1.00,1.03 \
34 sedge=0,0,0 \

6-1
 Build Model 4, Acquire a Line, Display Gathers, QC

35 6 xedge=1.80,2.00,2.20 \
36 zedge=0.98,1.05,1.03 \
37 sedge=0,0,0 \
38 7 xedge=7.0,7.5,8.0,8.5,9.0,9.50,10.0,11.0,12. \
39 zedge=0.5,0.6,0.7,0.8,0.9,0.91,0.88,0.72,0.5 \
40 sedge=0,0,0,0,0,0,0,0,0 \
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

41 8 xedge=7.0,7.5,8.0,8.50,9.00,9.50,10.0,10.5,11.0,11.5,12.0 \
42 zedge=0.5,0.7,0.9,1.02,1.15,1.25,1.25,1.20,1.15,1.08,1.00 \
43 sedge=0,0,0,0,0,0,0,0,0,0,0 \
44 9 xedge=-2,12 \
45 zedge=2,2 \
46 sedge=0,0 \
47 kedge=1,2,3,4,5,6,7,8,9 \
48 sfill=0.0,0.20,0,0,0.44,0,0 \
49 sfill=0.0,0.50,0,0,0.16,0,0 \
50 sfill=0.0,1.20,0,0,0.08,0,0 \
51 sfill=0.0,1.80,0,0,0.07,0,0 \
52 sfill=2.0,1.02,0,0,0.11,0,0 \
53 sfill=10.,1.00,0,0,0.09,0,0 > $modfile
54 ## x,z
55
56 # Create Encapsulated PostScript (EPS) image of model
57 spsplot < $modfile > $psfile \
58 gedge=0.5 gtri=2.0 gmin=0.0 gmax=5.0 \
59 title="Earth Model $num" \
60 labelz="Depth (km)" labelx="Distance (km)" \
61 wbox=5.25 hbox=0.75 dxnum=2.0 dznum=1.0
62
63 # Exit politely from shell
64 exit
65
This script can be divided into sets:
• System: Line 1 invokes the shell, line 5 turns on messages, and line 64 exits the
shell.
• Variables: Line 8 lets us vary the number of each run as we perfect the script (2a,
2b, etc.). This number becomes part of output file names. Line 11 assigns a name
to the output binary model file that is used on line 53. The output binary model
file is the prime reason for this script. Line 14 assigns a name to the output .eps
image file that is used on line 57.
• Bookkeeping: Line 17 removes a previous image file. This line is optional. On
some systems, the program crashes if a .eps file with the same name already
exists. Usually files are overwritten.
• Program trimodel: Lines 19-53 create the model. Notice that line 54 is a
comment. We use line 54 to remind us that the first two entries of line sfill are x-z
values. The very important sfill line is discussed below.
• Program spsplot: Lines 57-61 create the .eps image file.
Let’s examine program trimodel. Program trimodel fills the model with triangles of
(1/velocity)2. While (1/velocity) is called “slowness,” (1/velocity)2 is called “sloth.” A
sloth is a slow-moving tree-dwelling mammal found in Central and South America.
Program trimodel’s use here can be divided into five parts:
1. Line 19 defines the model dimensions.
2. The nine sets of xedge, zedge, and sedge triplets define layer boundaries
(interfaces). Each xedge, zedge, sedge triplet must have the same number of
values. For example, triplet 1 that defines the top of the model has two values to
define the flat layer:

6-2
 Build Model 4, Acquire a Line, Display Gathers, QC

20 1 xedge=-2,12 \
21 zedge=0,0 \
22 sedge=0,0 \
And, parameters xedge, zedge, sedge for interface two have eight values (as many
values as we thought necessary to define the interface):
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

23 2 xedge=-2.,0.00,2.0,4.0,6.0,8.00,10.0,12.0 \
24 zedge=0.3,0.32,0.3,0.6,0.2,0.25,0.25,0.25 \
25 sedge=0,0,0,0,0,0,0,0 \
Line sedge has only zeros because all layers are isotropic and homogeneous. Line
sedge would have non-zero values if we want a layer to have velocity gradients.
3. Use kedge, line 47, to list your interface numbers. This is important for your
acquisition script (the next section). An interface NOT listed here will NOT be
seen by a later acquisition script.
4. In this model that has nine boundaries, there are six layers (Figure 6.1): four
somewhat horizontal layers that extend across the model, a buried channel
(between interfaces 5 and 6) and a pinchout (between interfaces 7 and 8)
Therefore, there are six sfill lines, lines 48-53. These lines are written for
isotropic, homogenous layers, which is why most of the values are zero.

Figure 6.1: Model 4 is a 2-D slice of a marine environment (the top layer is water).
(TOP) The earth model. (MIDDLE) The simple model comprised of four layers; that is,
interfaces 1 (top), 2, 3, 4, and 9 (bottom). (BOTTOM) The model with the small channel
diffractor (interfaces 5 and 6) and the pinchout (interfaces 7 and 8). Interface 7 exactly
matches interface 3 from x = 7 km to x = 12 km. Interfaces 7 and 8 close (pinch out) at
x = 7 km.

6-3
 Build Model 4, Acquire a Line, Display Gathers, QC

Table 6.1: Line sfill variables

x z x0 z0 s00 ds/dx ds/dz
Each x-z pair of sfill specifies a point (any point) within a layer. Each sfill line
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

defines the sloth value that fills the layer. Remember, this model contains only
isotropic, homogenous layers.
Table 6.2: Values of sfill in Model 4
x z s00 velocity (1/¥sloth) Layer
(m) (m) (sloth) (m/s) boundaries
0.0 0.20 0.44 1508 1-2
0.0 0.50 0.16 2500 2-3
0.0 1.20 0.08 3536 3-4
0.0 1.80 0.07 3780 4-9
2.0 1.02 0.11 3015 5-6
10. 1.00 0.09 3333 7-8
Figure 6.2 shows that the sloth values generally decrease (P-wave velocity layer
values generally increase) with depth.

Figure 6.2: Model 4 layer velocities. Top: Sloth values. Bottom: Velocities in m/s.
To make interfaces 5 and 6 (below) create a layer, the two ends of the interfaces
intersect (see the bold x,z values below).
32 5 xedge=1.80,2.00,2.20 \
33 zedge=0.98,1.00,1.03 \

6-4
 Build Model 4, Acquire a Line, Display Gathers, QC

34 sedge=0,0,0 \
35 6 xedge=1.80,2.00,2.20 \
36 zedge=0.98,1.05,1.03 \
37 sedge=0,0,0 \
Interfaces 7 and 8 (below) have the same starting xedge (7.0) and zedge (0.5)
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

values (in bold below), ensuring that they form a pinchout. The interfaces form a
closed layer because the right ends intersect the right edge of the model.
38 7 xedge=7.0,7.5,8.0,8.5,9.0,9.50,10.0,11.0,12. \
39 zedge=0.5,0.6,0.7,0.8,0.9,0.91,0.88,0.72,0.5 \
40 sedge=0,0,0,0,0,0,0,0,0 \
41 8 xedge=7.0,7.5,8.0,8.50,9.00,9.50,10.0,10.5,11.0,11.5,12.0 \
42 zedge=0.5,0.7,0.9,1.02,1.15,1.25,1.25,1.20,1.15,1.08,1.00 \
43 sedge=0,0,0,0,0,0,0,0,0,0,0 \
5. Line 53 has the variable name of the output binary model file.
Program spsplot plots a triangulated sloth function as an encapsulated Postscript file.
So, we see that spsplot is designed to accompany trimodel! If you want to see the
triangles that are used to build the model, change gtri=2.0 to gtri=1.0.
On line 61, wbox=5.25 and hbox=0.75 ensure that the .eps image has the same
proportions as the actual model (line 19). It is important for us to specify the same width
and height in different programs within the same script so images created by different
programs exactly overlap. Also on line 61, dxnum and dznum specify the annotation
increments in the respective directions.

6.3 Acquire 2-D Line from Model 4: acq4.sh

In this section, we describe script acq4.sh that uses model4.dat to acquire a 2-D line
of seismic data. The survey layout:
1. 200 shots (line 33)
2. Shots are equally spaced at 50m intervals (line 37 and line 40)
3. 60 split-spread traces are recorded from each shot location (line 47)
4. Geophone spacing is 50m (line 51 and line 54)
5. Geophone offsets range from -1475 m to +1475 m (line 57)
6. Shot locations range from 0 km to 9.95 km (line 75)
7. Geophone locations range from -1.475 km to 11.425 km (line 75)
The final data set has 12000 traces (200 shots x 60 geophones per shot). The
generation of this data set took about thirty-six hours on a Sun UltraSPARC III with four
processors. For this reason, a copy of the output file is included on the CD that
accompanies this Primer.
1 #! /bin/sh
2 # File: acq4.sh
3 # Set messages on
4 ##set -x
5
6 # Assign values to variables
7 nangle=201 fangle=-65 langle=65 nt=501 dt=0.004
8
9 # Model
10 num=4
11 echo " --Model number = $num"
12
13 # Name input model file
14 inmodel=model$num.dat

6-5
 Build Model 4, Acquire a Line, Display Gathers, QC

15
16 # Name output seismic file
17 outseis=seis${num}.su
18
19 # Remove survey file
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

20 rm -f survey${num}.txt
21 # Name survey file
22 survey=survey${num}.txt
23
24 #=================================================
25 # Create the seismic traces with "triseis"
26 # i-loop = 200 source positions
27 # j-loop = 60 geophone positions (split-spread)
28 # per shot position
29 # k-loop = layers 2 through 8
30 # (do not shoot layers 1 and 9)
31
32 echo " --Begin looping over triseis."
33
34 i=0
35 while [ "$i" -ne "200" ]
36 do
37
38 fs=`bc -l <<-END
39 $i * 0.05
40 END`
41 sx=`bc -l <<-END
42 $i * 50
43 END`
44 fldr=`bc -l <<-END
45 $i + 1
46 END`
47
48 j=0
49 while [ "$j" -ne "60" ]
50 do
51
52 fg=`bc -l <<-END
53 $i * 0.05 + $j *0.05
54 END`
55 gx=`bc -l <<-END
56 $i * 50 + $j * 50 - 1475
57 END`
58 offset=`bc -l <<-END
59 $j * 50 - 1475
60 END`
61 tracl=`bc -l <<-END
62 $i * 60 + $j + 1
63 END`
64 tracf=`bc -l <<-END
65 $j + 1
66 END`
67
68 echo " Sx=$sx Gx=$gx fldr=$fldr Offset=$offset tracl=$tracl\
69 fs=$fs fg=$fg"
70 echo " Sx=$sx Gx=$gx fldr=$fldr Offset=$offset tracl=$tracl\
71 fs=$fs fg=$fg" >> $survey
72
73 k=2
74 while [ "$k" -ne "9" ]
75 do
76
77 triseis < $inmodel xs=0,9.95 xg=-1.475,11.425 zs=0,0 zg=0,0 \

6-6
 Build Model 4, Acquire a Line, Display Gathers, QC

78 nangle=$nangle fangle=$fangle langle=$langle \

79 kreflect=$k krecord=1 fpeak=40 lscale=0.5 \
80 ns=1 fs=$fs ng=1 fg=$fg nt=$nt dt=$dt |
81 suaddhead nt=$nt |
82 sushw key=dt,tracl,tracr,fldr,tracf,trid,offset,sx,gx \
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

83 a=4000,$tracl,$tracl,$fldr,$tracf,1,$offset,$sx,$gx >> temp$k

84
85 k=`expr $k + 1`
86
87 done
88 j=`expr $j + 1`
89
90 done
91 i=`expr $i + 1`
92
93 done
94
95 echo " --End looping over triseis."
96
97 #=================================================
98
99 # Sum contents of the temp files
100 echo " --Sum files."
101 susum temp2 temp3 > tempa
102 susum tempa temp4 > tempb
103 susum tempb temp5 > tempc
104 susum tempc temp6 > tempd
105 susum tempd temp7 > tempe
106 susum tempe temp8 > $outseis
107
108 # Remove temp files
109 echo " --Remove temp files."
110 rm -f temp*
111
112 # Report output file
113 echo " --Output file ** $outseis **"
114
115 # Exit politely from shell script
116 echo " --Finished!"
117 exit
118
This script writes messages to the screen to help us track progress:
11 echo " --Model number = $num"
32 echo " --Begin looping over triseis."
68 echo " Sx=$sx Gx=$gx fldr=$fldr Offset=$offset tracl=$tracl\
69 fs=$fs fg=$fg"
95 echo " --End looping over triseis."
100 echo " --Sum files."
109 echo " --Remove temp files."
113 echo " --Output file ** $outseis **"
116 echo " --Finished!"
Lines 68 and 69 write variable values to the screen. Notice that line 4 is commented out.
We find that once a script is perfected, the messages from line 4 interfere with reading
the values printed to the screen by lines 68 and 69.
Line 10 assigns values to variables nangle (the number of rays or angles that emanate
from the source), fangle (the first angle of the fan of rays that emanate from the source),
and langle (the last angle of the fan of rays that emanate from the source). The larger the
value of nangle, the longer it takes to produce the image. However, a complicated model
might require many rays so some will reflect to surface receivers.

6-7
 Build Model 4, Acquire a Line, Display Gathers, QC

Line 10 also assigns the number of time samples (nt=501) and the time sample
interval in seconds (dt=0.004). The output traces are 2 seconds long.
Note: Check the unit of the time sample interval in each SU program. The unit varies.
Lines 70 and 71 are almost the same as the previous two lines. However, where lines
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

68 and 69 write to the screen, lines 70 and 71 write values to a disk file. Although the
survey file for model 4 contains 12000 lines, it only occupies 1 Megabyte. Line 20
removes this file at the beginning of each run. If this line was not used, successive runs
would be appended (>>) to the end of previous runs. The survey file is named on line 22.
20 rm -f survey${num}.txt
22 survey=survey${num}.txt
70 echo " Sx=$sx Gx=$gx fldr=$fldr Offset=$offset tracl=$tracl\
71 fs=$fs fg=$fg" >> $survey
The core of the script is the three do-while loops (Section 2.3.13) over triseis,
suaddhead, and sushw. Each time these programs are used, triseis creates a seismic
trace, suaddhead creates a trace header on the trace, and sushw writes values to the trace
headers.
• The outermost i-loop goes over 200 source positions, 0 to 199 (lines 34 and 35).
• The j-loop goes over 60 geophone positions, 0 to 59 (lines 48 and 49).
• The innermost k-loop goes over the reflectors, 2 to 8 (lines 73 and 74).
Consider the innermost loop. The value of k signifies the reflector from which triseis
records a reflection (kreflect=$k, line 79) and the temp file in which the trace is stored
(temp$k, line 83.) Because k loops over reflectors 2 to 8, seven temp files are created.
(The top and base of the model are ignored.) Each temp data set has (in this case) 12000
traces. After all traces are created, lines 101 to 106 “sum” the seven data sets, trace-for-
trace.
The following pseudo-code of the do-while loops is presented to help us consider
how the variables are used. The program in which the variable is used is printed in bold
to the right on the assignment line. Variable k is underlined in lines T, Y, and CC to
remind us of its use as discussed in the preceding paragraph.
Variable fs (line E) is the source location relative to the start of the source surface.
The source surface is defined in line Y as variable xs. As i increments, the source moves
across the model. Both fs and sx are the source position. Variable fs is in model units,
kilometers, and is used by triseis; variable sx is an integer number of meters and is
computed for a header value.
Variable fg (line M) is the geophone location relative to the start of the geophone
surface. The geophone surface is defined in line Y as variable gx. As j increments, the
geophone moves across the model. Both fg and gx are the geophone position. Variable fg
is in model units, kilometers, and is used by triseis; variable gx is an integer number of
meters and is computed for a header value.
Variable fldr (line G) is a header value that identifies each shot gather. Variable tracl
(line P) numbers the traces continuously throughout the line. Variable tracf (line Q)
numbers the traces within each gather.
A i=0
B while i ne 200
C do

6-8
 Build Model 4, Acquire a Line, Display Gathers, QC

D
E fs = i * 0.05 triseis
F sx = i * 50 sushw
G fldr = i + 1 sushw
H
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

I j=0
J while j ne 60
K do
L
M fg = i * 0.05 + j *0.05 triseis
N gx = i * 50 + j * 50 - 1475 sushw
O offset = j * 50 - 1475 sushw
P tracl = i * 60 + j + 1 sushw
Q tracf = j + 1 sushw
R
S k=2
T while k ne 9
U do
V
W triseis < $inmodel xs=0,9.95 xg=-1.475,11.425 zs=0,0 zg=0,0 \
X nangle=$nangle fangle=$fangle langle=$langle \
Y kreflect=$k krecord=1 fpeak=40 lscale=0.5 \
Z ns=1 fs=$fs ng=1 fg=$fg nt=$nt dt=$dt |
AA suaddhead nt=$nt |
BB sushw key=dt,tracl,tracr,fldr,tracf,trid,offset,sx,gx \
CC a=4000,$tracl,$tracl,$fldr,$tracf,1,$offset,$sx,$gx >> temp$k
DD
EE k = k + 1
FF
GG done
HH j = j + 1
II
JJ done
KK i= i + 1
LL
MM done
Variable offset (line O) ranges from -1475 to +1475, incrementing by 50 meters. The
values of offset as it passes through zero offset are … -125, -75, -25, +25, +75, +125, etc.
This simple numbering scheme cleverly skips the geophysically impossible zero offset
acquisition position.
Let's use surange to examine the trace headers:
surange < seis4.su
The surange output is:
12000 traces:
tracl=(1,12000) tracr=(1,12000) fldr=(1,200) tracf=(1,60) trid=1
offset=(-1475,1475) sx=(0,9950) gx=(-1475,11425) ns=501 dt=4000

6.4 Model-Acquisition Checklist

The model script and the acquisition script must tie together.
• In general, always list all your interfaces in the model script’s kedge parameter;
for example, script model4.sh, line 47.
• Be sure your acquisition k-loop cycles through all the correct interfaces; for
example, script acq4.sh, lines 73-74.
You might have wondered why, in the model script, we always put the bottom
interface as the last interface (6th out of six in acq1.sh or 9th out of nine in

6-9
 Build Model 4, Acquire a Line, Display Gathers, QC

acq4.sh). We do it for the same reason we make the first model interface the top
interface. We do it because this approach simplifies designing the k-loop (script
acq4.sh, lines 73-87). By making the model’s bottom interface the last one (and
the top interface the first), the k-loop (triseis, suaddhead, sushw) begins at the
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

second interface and ends before the bottom interface.

On the other hand, if we put the bottom interface somewhere in the middle; say,
as interface five out of nine, this is not an overwhelming problem. We would
create two k-loops:
k=2
while [ "$k" -ne "5" ]
do
---
k=`expr $k + 1`
done
and
k=6
while [ "$k" -ne "10" ]
do
---
k=`expr $k + 1`
done
• Be sure you create and sum (susum) the correct number of temp files; for
example, script acq4.sh, lines 101-106.

6.5 Examine Shot Gathers

Now that we have seismic data (seis4.su), it is time to extract shot gathers for review.
Script showshot.sh (below):
• windows one gather from a seismic data set (line 14),
• makes a wiggle display of the seismic data (lines 17-19),
• converts the seismic data to a .eps file (lines 22-24), then
• writes the .eps image to output (the end of line 24).
The script expects two values on the command line, one for the model, another for the
shot number, as shown in lines 7 and 8.
This script assumes that all seismic data are in files of the same type of name as
shown underlined in line 14 (seis1.su, seis2.su, etc.), and in the same directory as the
script.
1 #! /bin/sh
2
3 # showshot.sh: Window one "field record" from file seis#.su
4 # where # represents the model number.
5 # Outputs: wiggle image of the shot gather
6 # .eps file of the shot gather
7 # Use: showshot.sh model shot
8 # Example: showshot.sh 3 20
9
10 # Set messages on
11 set -x
12
13 # Window one "field record" to a temporary file
14 suwind < seis$1.su key=fldr min=$2 max=$2 > temp$1$2.su
15

6-10
 Build Model 4, Acquire a Line, Display Gathers, QC

16 # Make wiggle plot

17 suxwigb < temp$1$2.su title="SP # $2 [$1]" key=offset \
18 label1=" Time (s)" label2="Offset (m)" \
19 x2beg=-1500 x2end=1500 perc=99 &
20
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

21 # Create .eps image of a shot gather

22 supswigp < temp$1$2.su title="SP # $2 [$1]" key=offset \
23 label1="Time (s)" label2="Offset (m)" \
24 x2beg=-1500 x2end=1500 perc=99 > shot$1$2.eps &
25
26 # Remove temporary gather
27 rm -f temp$1$2.su
28
29 # Exit politely from shell
30 exit
31
Entering
showshot.sh 4 45
commands showshot.sh to read file seis4.su, extract shot 45, and create file shot445.eps.
That is, the output .eps file name has the number of the model and the number of the shot,
as shown underlined in line 24. The title on the output images (lines 17 and 22) contains
the abbreviation SP (shot point), the number of the shot gather, and in square brackets,
the number of the seismic file (Figure 6.3).
This script does not output a .su file. A temporary .su file is created on line 14, is used
by suxwigb (line 17) and supswigp (line 22), and is removed by line 27. If you want a
permanent .su file:
• Change the word “temp” (lines 14, 17, and 22) to something more useful (like
“shot”).
• Comment out line 27 or remove lines 26 and 27.
• Add a comment line above line 5 describing the new output file.
• Remove the word “temporary” from comment line 13

Figure 6.3: Shot gather 45 from Model 4, shot almost directly over the channel.
The images below show selected shot gathers acquired over Model 4.

6-11
 Build Model 4, Acquire a Line, Display Gathers, QC

Offset (m) Offset (m) Offset (m)

-1000 0 1000 -1000 0 1000 -1000 0 1000
0 0 0
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

0.5 0.5 0.5

Time (s)

Time (s)

Time (s)
1.0 1.0 1.0

1.5 1.5 1.5

2.0 2.0 2.0

SP # 1 [4] SP # 10 [4] SP # 20 [4]

Offset (m) Offset (m) Offset (m)

-1000 0 1000 -1000 0 1000 -1000 0 1000
0 0 0

0.5 0.5 0.5

Time (s)

Time (s)

1.0 1.0 Time (s) 1.0

1.5 1.5 1.5

2.0 2.0 2.0

SP # 30 [4] SP # 40 [4] SP # 50 [4]

Offset (m) Offset (m) Offset (m)

-1000 0 1000 -1000 0 1000 -1000 0 1000
0 0 0

0.5 0.5 0.5

Time (s)

Time (s)

Time (s)

1.0 1.0 1.0

1.5 1.5 1.5

2.0 2.0 2.0

SP # 60 [4] SP # 70 [4] SP # 80 [4]

Figure 6.4.1: Shot gathers from Model 4. Numbers correspond to shot points.

6-12
 Build Model 4, Acquire a Line, Display Gathers, QC

Offset (m) Offset (m) Offset (m)

-1000 0 1000 -1000 0 1000 -1000 0 1000
0 0 0
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

0.5 0.5 0.5

Time (s)

Time (s)

Time (s)
1.0 1.0 1.0

1.5 1.5 1.5

2.0 2.0 2.0

SP # 90 [4] SP # 100 [4] SP # 110 [4]

Offset (m) Offset (m) Offset (m)

-1000 0 1000 -1000 0 1000 -1000 0 1000
0 0 0

0.5 0.5 0.5

Time (s)

Time (s)

1.0 1.0 Time (s) 1.0

1.5 1.5 1.5

2.0 2.0 2.0

SP # 120 [4] SP # 130 [4] SP # 140 [4]

Offset (m) Offset (m) Offset (m)

-1000 0 1000 -1000 0 1000 -1000 0 1000
0 0 0

0.5 0.5 0.5

Time (s)

Time (s)

Time (s)

1.0 1.0 1.0

1.5 1.5 1.5

2.0 2.0 2.0

SP # 150 [4] SP # 160 [4] SP # 170 [4]

Figure 6.4.2: Shot gathers from Model 4. Numbers correspond to shot points.

6-13
 Build Model 4, Acquire a Line, Display Gathers, QC

Offset (m) Offset (m) Offset (m)

-1000 0 1000 -1000 0 1000 -1000 0 1000
0 0 0
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

0.5 0.5 0.5

Time (s)

Time (s)

Time (s)
1.0 1.0 1.0

1.5 1.5 1.5

2.0 2.0 2.0

SP # 180 [4] SP # 190 [4] SP # 200 [4]

Figure 6.4.3: Shot gathers from Model 4. The channel creates the diffraction in SP 40
and SP 50. We first see the right side pinch-out in SP 150. Due to divergent ray paths, the
bottom layer does not return reflections in SP 150, SP 160, and SP 170.

6.6 Computer Note 9: Data Files Not in Script Directory

As stated in the previous section, script, showshot.sh, is written with the assumption
that our data files are in the same directory as the script. Let’s simplify and revise the
script to illustrate reading and writing data files that are not in the same directory as the
script. The scripts below read the seismic data output by the acquisition script, write a
selected shot gather to disk, and make a wiggle display of the single gather (read-write-
display).
Let’s start by learning two pointers:
. a single period refers to the current directory.
.. two periods together refers to the directory directly above the current directory.
If you enter:
ls -a
among other files listed you will see these two pointers. Every Unix directory is
connected to the others by these pointers. You can think of the single period as “here”
and the double period as “up one directory.” We can use these pointers instead of using
full path names in scripts (and line commands).

6.6.1 Data Directory below Script Directory — Full Path Name

We sometimes find it convenient to develop our scripts in a directory separate from
our data. In this case, it is easy to refer to a data directory directly below the current
directory. Before showing the easy way, we show the full path name way.
1 #! /bin/sh
2
3 # shotrwd1.sh: Window one "field record" from file seis#.su
4 # where # represents the model number.
5 # Outputs: .su file of the shot gather
6 # wiggle image of the shot gather
7 # Use: shotrwd1.sh model shot
8 # Example: shotrwd1.sh 3 20

6-14
 Build Model 4, Acquire a Line, Display Gathers, QC

9
10 # Set messages on
11 set -x
12
13 # Window one "field record"
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

14 suwind < /home/forel/suscripts/data/seis$1.su \

15 key=fldr min=$2 max=$2 \
16 > /home/forel/suscripts/data/shot$1$2.su
17
18 # Make wiggle plot
19 suxwigb < /home/forel/suscripts/data/shot$1$2.su \
20 title="SP # $2 [$1]" key=offset \
21 label1=" Time (s)" label2="Offset (m)" \
22 x2beg=-1500 x2end=1500 perc=99 &
23
24 # Exit politely from shell
25 exit
26
Entering
shotrwd1.sh 4 130
commands shotrwd1.sh to read file seis4.su that is in subdirectory data (line 14), extract
shot 130, write file shot4130.su to subdirectory data (line 16), read the single gather from
directory data (line 19), and display field record 130 as a wiggle plot.

6.6.2 Data Directory below Script Directory — Short Path Name

Assume we are in directory /home/forel/suscripts. Let’s rewrite the script using short
path names. Remember to change the script’s file name and the internal references to this
file name (lines 3, 7, and 8).
1 #! /bin/sh
2
3 # shotrwd2.sh: Window one "field record" from file seis#.su
4 # where # represents the model number.
5 # Outputs: .su file of the shot gather
6 # wiggle image of the shot gather
7 # Use: shotrwd2.sh model shot
8 # Example: shotrwd2.sh 3 20
9
10 # Set messages on
11 set -x
12
13 # Window one "field record"
14 suwind < ./data/seis$1.su \
15 key=fldr min=$2 max=$2 \
16 > ./data/shot$1$2.su
17
18 # Make wiggle plot
19 suxwigb < ./data/shot$1$2.su \
20 title="SP # $2 [$1]" key=offset \
21 label1=" Time (s)" label2="Offset (m)" \
22 x2beg=-1500 x2end=1500 perc=99 &
23
24 # Exit politely from shell
25 exit
26
Entering
shotrwd2.sh 4 150

6-15
 Build Model 4, Acquire a Line, Display Gathers, QC

commands shotrwd2.sh to start here (./), go to subdirectory data, read file seis4.su (line
14), extract shot 130, write file shot4150.su to subdirectory data that is below the current
directory (./) (line 16). Then (line 19) says, from here (./), go to subdirectory data, read
the single gather of field record 150 and display it as a wiggle plot.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Note: You do not need to use./ in the above script. The following script is just as
effective.
1 #! /bin/sh
2
3 # shotrwd3.sh: Window one "field record" from file seis#.su
4 # where # represents the model number.
5 # Outputs: .su file of the shot gather
6 # wiggle image of the shot gather
7 # Use: shotrwd3.sh model shot
8 # Example: shotrwd3.sh 3 20
9
10 # Set messages on
11 set -x
12
13 # Window one "field record"
14 suwind < data/seis$1.su \
15 key=fldr min=$2 max=$2 \
16 > data/shot$1$2.su
17
18 # Make wiggle plot
19 suxwigb < data/shot$1$2.su \
20 title="SP # $2 [$1]" key=offset \
21 label1=" Time (s)" label2="Offset (m)" \
22 x2beg=-1500 x2end=1500 perc=99 &
23
24 # Exit politely from shell
25 exit
26

6.6.3 Data Directory at the same Level as Script Directory

Suppose you have a directory worksu and under worksu are two directories: scripts
and data. Our problem is to make a script that is in directory “scripts” read the data
directory without using full path names. Here is the solution:
1 #! /bin/sh
2
3 # shotrwd4.sh: Window one "field record" from file seis#.su
4 # where # represents the model number.
5 # Outputs: .su file of the shot gather
6 # wiggle image of the shot gather
7 # Use: shotrwd4.sh model shot
8 # Example: shotrwd4.sh 3 20
9
10 # Set messages on
11 set -x
12
13 # Window one "field record"
14 suwind < ../data/seis$1.su \
15 key=fldr min=$2 max=$2 \
16 > ../data/shot$1$2.su
17
18 # Make wiggle plot
19 suxwigb < ../data/shot$1$2.su \
20 title="SP # $2 [$1]" key=offset \
21 label1=" Time (s)" label2="Offset (m)" \

6-16
 Build Model 4, Acquire a Line, Display Gathers, QC

22 x2beg=-1500 x2end=1500 perc=99 &

23
24 # Exit politely from shell
25 exit
26
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Our script is in directory scripts. On line 14, “../” directs the system to go up one
directory (to directory worksu). The rest of the path name specifies the data file going
down from directory worksu.
You can use “../” repeatedly to climb up the directory structure. For example,
suwind < ../../worksu/data/seis$1.su \
tells the system to go from directory scripts up to worksu, up to forel, down to worksu,
down to data. Don’t specify file names going up.

6.7 Quality Control: The Survey File

We can run an acquisition script quickly to learn if the computed header values are
the ones we want. We do this by commenting out the SU programs, running the
acquisition script, then examining the output survey file. Since the SU programs are not
used, the script executes quickly. The acquisition script that took thirty-six hours takes
only half an hour when we comment out lines 77-83.
77 ## triseis < $inmodel xs=0,9.95 xg=-1.475,11.425 zs=0,0 zg=0,0 \
78 ## nangle=$nangle fangle=$fangle langle=$langle \
79 ## kreflect=$k krecord=1 fpeak=40 lscale=0.5 \
80 ## ns=1 fs=$fs ng=1 fg=$fg nt=$nt dt=$dt |
81 ## suaddhead nt=$nt |
82 ## sushw key=dt,tracl,tracr,fldr,tracf,trid,offset,sx,gx \
83 ## a=4000,$tracl,$tracl,$fldr,$tracf,1,$offset,$sx,$gx >> temp$k

100 ##echo " --Sum files."

101 ##susum temp2 temp3 > tempa
102 ##susum tempa temp4 > tempb
103 ##susum tempb temp5 > tempc
104 ##susum tempc temp6 > tempd
105 ##susum tempd temp7 > tempe
106 ##susum tempe temp8 > $outseis

109 ##echo " --Remove temp files."

110 ##rm -f temp*

113 ##echo " --Output file ** $outseis **"

Since no traces are created (no temp files), the script would crash if it tried to execute
lines 101-106; so, we comment them out. If executed, line 110 would not make the script
crash, but there is nothing wrong with commenting it out. We comment out lines 100,
109, and 113 because we do not want to read false information.
Notice that these lines are commented out with two “#.” We know only one is
necessary. We use two so we see where we intend comments to be temporary.
Below are the first ten lines and the last ten lines of survey4.txt.
Sx=0 Gx=-1475 fldr=1 Offset=-1475 tracl=1 fs=0 fg=0
Sx=0 Gx=-1425 fldr=1 Offset=-1425 tracl=2 fs=0 fg=.05
Sx=0 Gx=-1375 fldr=1 Offset=-1375 tracl=3 fs=0 fg=.10
Sx=0 Gx=-1325 fldr=1 Offset=-1325 tracl=4 fs=0 fg=.15
Sx=0 Gx=-1275 fldr=1 Offset=-1275 tracl=5 fs=0 fg=.20
Sx=0 Gx=-1225 fldr=1 Offset=-1225 tracl=6 fs=0 fg=.25
Sx=0 Gx=-1175 fldr=1 Offset=-1175 tracl=7 fs=0 fg=.30

6-17
 Build Model 4, Acquire a Line, Display Gathers, QC

Sx=0 Gx=-1125 fldr=1 Offset=-1125 tracl=8 fs=0 fg=.35

Sx=0 Gx=-1075 fldr=1 Offset=-1075 tracl=9 fs=0 fg=.40
Sx=0 Gx=-1025 fldr=1 Offset=-1025 tracl=10 fs=0 fg=.45

Sx=9950 Gx=10975 fldr=200 Offset=1025 tracl=11991 fs=9.95 fg=12.45

Sx=9950 Gx=11025 fldr=200 Offset=1075 tracl=11992 fs=9.95 fg=12.50
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Sx=9950 Gx=11075 fldr=200 Offset=1125 tracl=11993 fs=9.95 fg=12.55

Sx=9950 Gx=11125 fldr=200 Offset=1175 tracl=11994 fs=9.95 fg=12.60
Sx=9950 Gx=11175 fldr=200 Offset=1225 tracl=11995 fs=9.95 fg=12.65
Sx=9950 Gx=11225 fldr=200 Offset=1275 tracl=11996 fs=9.95 fg=12.70
Sx=9950 Gx=11275 fldr=200 Offset=1325 tracl=11997 fs=9.95 fg=12.75
Sx=9950 Gx=11325 fldr=200 Offset=1375 tracl=11998 fs=9.95 fg=12.80
Sx=9950 Gx=11375 fldr=200 Offset=1425 tracl=11999 fs=9.95 fg=12.85
Sx=9950 Gx=11425 fldr=200 Offset=1475 tracl=12000 fs=9.95 fg=12.90
Comparing the output to line 77 (triseis), observe that:
• the first Sx value corresponds to the first xs value of line 77 (0),
• the last Sx value corresponds to the last xs value of line 77 (9.95),
• the first Gx value corresponds to the first xg value of line 77 (-1.475), and
• the last Gx value corresponds to the last xg value of line 77 (11.425).

6.8 Quality Control: Acquire and Display Specific Shot Gathers

Before shooting the entire line, you might want to acquire single shot gathers to
preview seismic data and header values. It is surprisingly simple to modify the
acquisition script to do this.
• Copy script acq4.sh to a new file and document that change in line 2 of the new
script.
2 # File: acq4shot.sh
• Change the names of the output files. Line 17 contains the name of the output
seismic file. Lines 20, 22, and 71 contain the name of the survey file.
17 outseis=seis${num}shot.su
20 rm -f survey${num}shot.txt
22 survey=survey${num}shot.txt
70 echo " Sx=$sx Gx=$gx fldr=$fldr Offset=$offset tracl=$tracl\
71 fs=$fs fg=$fg" >> $survey
• Most important, specify the shot gather using lines 34 and 35. Here, we acquire
gather 50:
34 i=49
35 while [ "$i" -ne "50" ]
No other changes are necessary.
A similar change acquires several shot gathers together. The following two lines
acquire gathers 51-55.
34 i=50
35 while [ "$i" -ne "55" ]
If you do not change output file names, you run the risk of overwriting valuable data
during future runs.
The script below, showshotB.sh, is designed to accompany script acq4shot.sh. Lines 3
and 14 differ from script showshot.sh in specifying the new input file.
1 #! /bin/sh
2
3 # showshotB.sh: Window one "field record" from file seis#shot.su
4 # where # represents the model number.
5 # Outputs: wiggle image of the shot gather

6-18
 Build Model 4, Acquire a Line, Display Gathers, QC

6 # .eps file of the shot gather

7 # Use: showshotB.sh model shot
8 # Example: showshotB.sh 3 20
9
10 # Set messages on
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

11 set -x
12
13 # Window one temporary "field record"
14 suwind < seis$1shot.su key=fldr min=$2 max=$2 > temp$1$2.su
15
16 # Make wiggle plot
17 suxwigb < temp$1$2.su title="SP # $2 [$1]" key=offset \
18 label1=" Time (s)" label2="Offset (m)" \
19 x2beg=-1500 x2end=1500 perc=99 &
20
21 # Create .eps image of a shot gather
22 supswigp < temp$1$2.su title="SP # $2 [$1]" key=offset \
23 label1="Time (s)" label2="Offset (m)" \
24 x2beg=-1500 x2end=1500 perc=99 > shot$1$2.eps &
25
26 # Remove temporary gather
27 rm -f temp$1$2.su
28
29 # Exit politely from shell
30 exit
31
Entering
showshotB.sh 4 50
commands showshotB.sh to read file seis4shot.su, extract shot 50, and create file
shot450.eps. That is, the output .eps file name has the number of the model and the
number of the shot, as highlighted in line 24. The title on the output images (lines 17 and
22) contains the abbreviation SP (shot point), the number of the shot gather, and in square
brackets, the number of the seismic file (Figure 6.3).
This script does not output a .su file. A temporary .su file is created on line 14, is used
by suxwigb (line 17) and supswigp (line 22), and is removed by line 27. If you want a
permanent .su file, follow the instructions in section 6.5.

6-19
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

This page has been intentionally left blank

 Model 4: Sort, Velocity Analysis

7. Model 4: Sort, Velocity Analysis

7.1 Introduction
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

From the previous chapter, we have a 2-D line of seismic data. In this chapter, we:
• sort the shot gathers to common midpoint (CMP) gathers and
• perform velocity analysis on selected CMP gathers.
Because our synthetic data are noise-free, we can proceed quickly from shot gathers to
migration. If we wanted to, we could use suaddnoise to add noise to the data; this is
often useful. You should consider adding some noise to the data set, then processing it
using the examples we provide in this chapter and the next.
Let’s discuss “common-depth-point” (CDP) and “common midpoint” (CMP) (Sheriff,
2002). We try to avoid the term CDP because there is no common (same) point at the
reflector if the reflector dips. On the other hand, common-midpoints almost always exist
because they are defined by the geometric midpoint between sources and receivers.
Because SU does not have a cmp key, we reluctantly use the cdp key.

Figure 7.1: Left: The common-depth-point (CDP) and the common-midpoint (CMP) are
the same because the reflector is flat. Right: The reflections in this CMP gather do not
share a CDP because the reflector dips.

7.2 Sort from Shot Gathers to CMP Gathers

Script sort2cmp.sh does two jobs:
1. We use program suchw (Change Header Word) to create header (key) cdp and
assign values to it. Using the equation below, values of cdp (key1) are computed
from gx (key2) and sx (key3).
key1 = ( a + b*key2 + c*key3 ) / d
Scalars a, b, c, and d need to be determined by sketching the geometry. Our
geometry dictates a = 1525, b =1, c = 1, and d = 50. Therefore,
cdp = ( 1525 + 1*gx + 1*sx ) / 50
2. We use program susort to sort the traces based on primary sort key cdp and
secondary sort key offset. A secondary sort is the order within the primary sort
entity.
1 #! /bin/sh
2 # sort2cmp.sh
3

7-1
 Model 4: Sort, Velocity Analysis

4 # Compute CMP header

5 suchw < seis4.su key1=cdp key2=gx key3=sx \
6 a=1525 b=1 c=1 d=50 |
7
8 # Sort data to CMP gathers
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

9 # file 1 2
10 susort > cmp4.su cdp offset
11
12 # Exit politely from shell
13 exit 0
14
Let’s try to understand how the first part of the script works. Typically, when we
assign cdp values, we want the output CMP gathers to increment by “1.”
Recall that receiver positions increment by 50 meters and the source positions also
increment by 50 meters. In the equation above, scalar “a” is set to 50 meters greater than
the largest positive offset. Scalar “d” is chosen to normalize the values computed in the
numerator so cdp values increment simply.
Table 7.1 shows headers sx and gx of the first and last traces of the first ten shots, as
well as the computed cdp values. You can see that cdp values slowly increase in a regular
way.
Table 7.1: cdp values computed from headers sx and gx
cdp sx gx
tracf = 1 1 0 -1475
Shot 1
tracf = 60 60 0 1475
tracf = 1 3 50 -1425
Shot 2
tracf = 60 62 50 1525
tracf = 1 5 100 -1375
Shot 3
tracf = 60 64 100 1575
tracf = 1 7 150 -1325
Shot 4
tracf = 60 66 150 1625
tracf = 1 9 200 -1275
Shot 5
tracf = 60 68 200 1675
tracf = 1 11 250 -1225
Shot 6
tracf = 60 70 250 1725
tracf = 1 13 300 -1175
Shot 7
tracf = 60 72 300 1775
tracf = 1 15 350 -1125
Shot 8
tracf = 60 74 350 1825
tracf = 1 17 400 -1075
Shot 9
tracf = 60 76 400 1875
tracf = 1 19 450 -1025
Shot 10
tracf = 60 78 450 1925
Table 7.2 shows headers sx and gx of the first and last traces of the last ten shots, as
well as the computed cdp values. You can see that cdp values slowly increase. Also,
notice that the last cdp value is 458 – recall that there are 176 shot gathers!
Table 7.2: cdp values computed from headers sx and gx
cdp sx gx
tracf = 1 381 9500 8025
Shot 191
tracf = 60 440 9500 10975
tracf = 1 383 9550 8075
Shot 192
tracf = 60 442 9550 11025

7-2
 Model 4: Sort, Velocity Analysis

tracf = 1 385 9600 8125

Shot 193
tracf = 60 444 9600 11075
tracf = 1 387 9650 8175
Shot 194
tracf = 60 446 9650 11125
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

tracf = 1 389 9700 8225

Shot 195
tracf = 60 448 9700 11175
tracf = 1 391 9750 8275
Shot 196
tracf = 60 450 9750 11225
tracf = 1 393 9800 8325
Shot 197
tracf = 60 452 9800 11275
tracf = 1 395 9850 8375
Shot 198
tracf = 60 454 9850 11325
tracf = 1 397 9900 8425
Shot 199
tracf = 60 456 9900 11375
tracf = 1 399 9950 8475
Shot 200
tracf = 60 458 9950 11425
Below is the surange output of file cmp4.su.
surange < cmp4.su
12000 traces:
tracl=(1,12000) tracr=(1,12000) fldr=(1,200) tracf=(1,60) cdp=(1,458)
trid=1 offset=(-1475,1475) sx=(0,9950) gx=(-1475,11425) ns=501
dt=4000
Below is the output of the first ten lines of command:
sugethw < cmp4.su key=cdp,offset,fldr
cdp=1 offset=-1475 fldr=1
cdp=2 offset=-1425 fldr=1
cdp=3 offset=-1475 fldr=2
cdp=3 offset=-1375 fldr=1
cdp=4 offset=-1425 fldr=2
cdp=4 offset=-1325 fldr=1
cdp=5 offset=-1475 fldr=3
cdp=5 offset=-1375 fldr=2
cdp=5 offset=-1275 fldr=1
cdp=6 offset=-1425 fldr=3
This list helps us see that the data are in cdp order (key fldr is no longer the primary
sort order) with secondary key of offset. We also see that fold slowly increases: there is
one line for cdp = 1 and one line for cdp = 2; there are two lines for cdp = 3 and two lines
for cdp = 4; there are three lines for cdp = 5, etc.
The output of surange shows that the cdp values range from “1” to “458” (underlined
above), confirming our earlier calculations.

7.3 Extract CMP Gathers

Now that we have a line of CMP gathers (cmp4.su), let’s extract individual CMPs for
view and for the next step: velocity analysis. Script showcmp.sh (below):
• windows one gather from the line and write it to disk (line 20),
• makes a wiggle display of the .su CMP gather (lines 22-24),
• converts the .su CMP gather to a .eps file (lines 26-28), then
• writes the .eps image to disk (the end of line 28).
The script expects two values on the command line, one for the model, another for the
CMP number, as shown in lines 6 and 7.

7-3
 Model 4: Sort, Velocity Analysis

This script assumes that all seismic data are in files of the same type of name as
underlined in line 20 (cmp1.su, cmp2.su, etc.), and in the same directory as the script.
1 #! /bin/sh
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

2
3 # File: showcmp.sh: Window one CMP from file seis#.su
4 # where # represents the model number.
5 # Specify input model number and output CMP
6 # Input: CMP file
7 # Outputs: CMP gather in .su format
8 # wiggle image of the CMP gather
9 # .eps file of the CMP gather
10 # Use: showcmp.sh [model] [cmp]
11 # Example: showcmp.sh 4 20
12
13 # Set messages on
14 ##set -x
15
16 # Display a CMP gather between 1 and 458
17 if [ $2 -le 458 ]; then
18 if [ $2 -ge 1 ]; then
19
20 suwind < cmp$1.su key=cdp min=$2 max=$2 > cmp$1$2.su
21
22 suxwigb < cmp$1$2.su title="CMP # $2 [$1]" key=offset \
23 label1=" Time (s)" label2="Offset (m)" \
24 perc=99 &
25
26 supswigp < cmp$1$2.su title="CMP # $2 [$1]" key=offset \
27 label2="Time (s)" label2="Offset (m)" \
28 x2beg=-1500 x2end=1500 perc=99 > cmp$1$2.eps
29
30 exit
31
32 fi
33 fi
34
35 echo usage: showcmp.sh [model number] [CMP between 1 and 458]
36
37 # Exit politely from shell
38 exit
39

7.4 Computer Note 10: If Blocks

The script above (showcmp.sh) uses two nested if blocks (Section 2.3.13). Script
showshot.sh (Section 6.5) did not use if blocks, and they are not necessary here, but we
take this opportunity to illustrate them.
The if blocks ensure that we input a CMP value (variable $2) that is within the range
of actual values. Line 15 requires that the value of $2 is 458 or less; line 16 requires that
the value of $2 is 1 or greater.
If an invalid value is input for $2, for example:
showcmp.sh 4 900
the “echo” message of line 33 is written to the screen:
usage: showcmp.sh [model number] [CMP between 1 and 458]
If a valid value for $2 is input, the “echo” message is skipped.

7-4
 Model 4: Sort, Velocity Analysis

If you think there is a problem with the script or how you are using the script,
uncomment line 14.

7.5 Fold
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

After sorting to CMP-order, we have 458 gathers. But, the first gathers and the last
gathers of the line are not full fold. See Figure 7.2, below. (We describe the program that
generated the data for Figure 7.2, sukeycount, in Appendix C.) Upon inspection, we find
that the first full-fold CMP gather is CMP 60 and the last full fold CMP gather is CMP
400. Full fold is 30 traces, in contrast to the shot gather full fold of 60 traces. This is an
expected outcome of sorting based on the acquisition geometry (see Sheriff, 2002:
Stacking Chart).
To make reasonably detailed velocity analysis at a regular interval along the line, we
choose to analyze eighteen full-fold CMPs: 60, 80, 100, …, 360, 380, 400.

Figure 7.2: Fold values for data set cmp4.su.

7.6 Stacking Velocity Analysis

There are several types of velocities in seismic reflection work: interval velocity,
root-mean-square (rms) velocity, stacking velocity, and migration velocity are some. At
this stage of processing, we want to determine velocities that allow us to stack each CMP
gather into a single trace. To reinforce primary reflections in the (following) stacking
process, we want to determine a time-velocity (t-v) function that flattens reflections in
each CMP that we analyze.
The following sections describe two tools that we later incorporate into an interactive
velocity analysis (iva) script: the semblance plot and constant velocity stacks.

7.6.1 Prepare a Shot Gather for Velocity Analysis

To demonstrate how we prepare a real (non-synthetic) gather for semblance, we
download file ozdata.14 from the CWP ftp site. Following our file naming convention,
we changed the file name to oz14.su: This file has 6.3 seconds of data, more than we need
for our demonstration (see Figure 7.3 below).
Below is the surange output of file oz14.su.

7-5
 Model 4: Sort, Velocity Analysis

surange < oz14.su

48 traces:
tracl=(1,48) tracr=(1,48) fldr=10014 tracf=(1,48) cdp=(14,61)
cdpt=1 trid=1 nvs=4 nhs=1 duse=1
scalel=1 scalco=1 counit=1 delrt=4 muts=4
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

ns=1575 dt=4000

Figure 7.3: Original data file oz14.su has 48 traces and is 6.3 seconds long.
The following script, oz14prep.sh, prepares oz14.su for velocity analysis.
1 #! /bin/sh
2
3 # File: oz14prep.sh: Prepare oz14.su for velocity analysis
4 # Input (1): shot gather
5 # Output (1): modified CMP gather
6 # Use: oz14prep.sh
7
8 # Set messages on
9 set -x
10
11 # Name data sets
12 indata=oz14.su
13 outdata=oz14h.su
14
15 # Use only first 4 seconds of data
16 suwind < $indata tmax=4 |
17
18 # Gain
19 sugain tpow=2 |
20
21 # Convert from shot gather to CMP gather
22 # by making all cdp values = 14
23 suchw key1=cdp a=14 b=0 |
24
25 # Add offset values to offset key
26 suchw key1=offset a=11250 b=-220 key2=tracf |
27
28 # Mute

7-6
 Model 4: Sort, Velocity Analysis

29 sumute key=tracf \
30 xmute=1,46,48 \
31 tmute=2.250,0.500,0.004 \
32 > $outdata
33
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

34 # Plot
35 suxwigb < $outdata key=offset perc=90 &
36
37 # Exit politely from shell
38 exit
39
In oz14prep.sh, the data are prepared in the following ways:
• Line 16 windows the first 4 seconds of data.
• Line 19 applies a classic power of 2 time-varying gain.
• Line 23 “effectively” converts the shot gather to a CMP gather by making all
values of key cdp the same. (See Section 3.7.)
• Line 26 puts values into key offset according to either of Yilmaz’ books: Table 1-
8 of Seismic Data Processing (1987) or Table 1-13 of Seismic Data Analysis
(2001): the nearest offset trace is 69 feet from the source and trace spacing is 220
feet. Since the first trace is the farthest offset trace, parameter a has the farthest
offset and parameter b is used to subtract the trace spacing from parameter a.
• Lines 29-31 mute the refractions. Figure 7.4, below left, shows the data after
oz14prep.sh. For comparison, the right side of Figure 7.4 shows the data after all
the processing of oz14prep.sh, but without the mute.

Figure 7.4: Left: Output of script oz14prep.sh. Right: Same as Left, but without mute.

7-7
 Model 4: Sort, Velocity Analysis

7.6.2 Semblance Contour Plot

Velocity analysis, a necessary prelude to NMO correction, yields the stacking
velocities. NMO correction is also called flattening the data. Once CMP gathers are
flattened, they can be stacked. Stacking attenuates random noise and multiples. Stacking
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

yields a zero-offset section (neglecting elastic or amplitude-versus-offset effects) that is

ready to be migrated. (We ignore dip moveout correction (DMO).) After migration, the
data are interpreted for geological meaning.
We would like to decrease the time it takes to do velocity analysis. One way is to
limit the range of velocities that we consider to be in the data. The range of velocities is
usually limited by intelligent guessing. We know that a hyperbola with a lot of curvature
represents a low velocity and a hyperbola with little curvature represents a high velocity.
Figure 7.5 (below) is CMP 60 from model 4. There are three obvious reflectors on this
gather. We know that the upper red curve represents a velocity less than any reflector.
We also know that the lower red curve represents a velocity greater than any reflector.
We can use this knowledge to limit the number of velocities to compute. Also, when we
know the top layer is water, as in Model 4, the slowest velocity is 1500 m/s or 4920 f/s.

Figure 7.5: CMP 60 from Model 4. The lower and upper lines (hyperbolas) are the
lower and upper velocities necessary for velocity analysis.
(Remember that color figures and those requiring detail are included as TIFF files on
a CD accompanying this book, as listed in Appendix E.)

7-8
 Model 4: Sort, Velocity Analysis

Unfortunately, it is not yet possible to push a button and have the computer figure out
the velocities that flatten the primary reflections. Remember, real seismic data have
multiples and noise that interfere with moveout analysis. The synthetic gather in Figure
7.5 also has noise, but not so much that we can't easily see the reflectors.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

When we think we have a reasonable velocity range, we create a plot of a velan

(VELocity ANalysis) of the CMP. SU program suvelan computes stacking velocities.
We send (pipe) the output of suvelan to suxcontour to make a plot. Then, we pick time-
velocity (t-v) pairs from the velan.
Below is script oz14velan.sh that creates a velan from CMP oz14h.su. File oz14h.su
has noise and multiples, so velocity analysis is not easy.
1 #! /bin/sh
2
3 # File: oz14velan.sh: Create semblance display of one CMP
4 # Input (1): One cdp gather
5 # Output (1): Velan (contour) plot
6 # Use: oz14velan.sh
7
8 # Set messages on
9 set -x
10
11 # Input data
12 indata=oz14h.su
13
14 # Important processing variables
15 # nv = number of vels, dv = delta vel, fv = first vel
16 nv=160
17 dv=50
18 fv=4000
19
20 # Do the work
21
22 suvelan < $indata nv=$nv dv=$dv fv=$fv smute=1.5 |
23 suxcontour bclip=0.5 wclip=0.0 f2=$fv d2=$dv \
24 label1=" Time (s)" label2="Velocity" \
25 title="Semblance plot for CMP $indata" \
26 windowtitle="Semblance: $indata" \
27 grid1=solid grid2=solid cmap=default \
28 nc=7 nlabelc=0 &
29
30 # Exit politely from shell
31 exit
32
Figure 7.6 is the velan generated by script oz14velan.sh. The core mathematics in
suvelan is called semblance. The velan display is actually a plot of semblance.
Semblance is defined in the suvelan selfdoc.
Here, we describe the velan display without mathematics. The input CMP has offset
values in the trace headers. The time of each data sample is also known. Program suvelan
goes through the times of a fictitious zero-offset trace. At each time on this zero-offset
trace, it applies NMO and stack along a series of hyperbolas – the hyperbolas that
correspond to velocities from 4000 to 12000 ( = (50 x 160) + 4000 ) f/s. It does this in
increments of 50 (the value of dv).
On velans, the time axis specifies zero-offset time. The display is a contour plot of
amplitude reinforcement from applying moveout and stack using the range of velocities.

7-9
 Model 4: Sort, Velocity Analysis

Here we use suxcontour to show black contour lines on a white background. We could
also use suximage to create color-area displays on a computer monitor.
Parameter nc is a sort of volume control – by varying the number of contour levels,
we coarsen or fine-tune the output of the semblance calculation.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 7.6: Semblance plot with t-v picks.

In Figure 7.6, solid dots show where t-v picks were made. The vertical line labeled
“Multiples” has a single velocity, 6800 f/s. Although events in the left side of Figure 7.7
are nicely flattened, this is misleading. We want multiples to stack out (attenuate), not
reinforce. The NMO plot in the right side of Figure 7.7, the result of velan picks for
primaries, shows poor event alignment below approximately 2.5 seconds. We see in
Figure 7.7 that multiples cover up primaries. The best we can do is pick t-v pairs that will
attenuate the multiples during stack. We do this so that after stack, the primaries will be
easier to see (interpret).
Note: The mutes are different on the two data sets in Figure 7.7 because sunmo
parameter smute affected the two moveout curves differently.
Table 7.3: Time-Velocity picks on CMP oz14h.su
Multipes Primaries
tnmo=0.010,1.280,3.980 tnmo=0.020,1.590,2.520,3.97
vnmo=5000,6800,6800 vnmo=5000,7180,7980,10200

7-10
 Model 4: Sort, Velocity Analysis
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 7.7: Left: oz14h.su after NMO for multiples. Right: oz14h.su after NMO for
primaries.

7.6.3 Constant Velocity Stacks

Another velocity analysis plot is constant velocity stacks (CVS). Figure 7.8 is the
result of making a CVS from CMP 60 of Model 4 (Figure 7.5). Figure 7.8 shows ten CVS
panels. Each panel has 11 traces — always an odd number. (Individual traces within
panels are difficult to see because we used suximage instead of suxwigb.)
The middle trace of each panel is CMP 60. The first CVS panel, the first 11 traces, is
the result of applying NMO and stack to CMPs 55 to 65. This is also true for the next
nine panels. (CVS plots cannot be made for a single CMP; only for a line of CMPs.) The
NMO velocity for the first CVS panel is 1200 m/s. The NMO velocity for the second
CVS panel is 1500 m/s. To make Figure 7.8, NMO and stack was done 110 times!
Each panel gives a simplified view of the final stacked section. The more continuity
the events show and the higher their amplitudes, the better will be the final stacked result.
Because each panel uses only one velocity, the best picks over time usually do not fall
within one single panel. (Note: This can be different for multi velocity stacks (MVS).
Each panel of an MVS represents a time-varying velocity function. MVS are common in
the industry today, but the idea behind CVS is not much different and it is easier to create
in SU.) The first reflector has the highest amplitude (is best focused) near 1500 m/s, the
second reflector has the highest amplitude near 2700 m/s, and the third reflector has the
highest amplitude near 3000 m/s.
The arrows in Figure 7.8 emphasize that when we interpret the best velocity from
each panel, we use the velocity above the center trace of each panel. The other traces in

7-11
 Model 4: Sort, Velocity Analysis

the panel provide visual emphasis to the best NMO velocity, but the center CMP of each
panel is the one we study.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 7.8: A constant velocity stack (CVS) of 10 panels made from CMP 60, Model 4.
Also, remember that the velocity spacing between panels (here) is 300 m/s. We
cannot determine a precise stacking velocity from the CVS plot, but it is valuable when
used with the velan. We will show this in a later section when we combine the original
CMP display, the velan, and the CVS plot in an interactive velocity analysis script.

7.6.4 Quality Control (QC) after Picking

The semblance plot and the CVS are tools we use to decide which stacking velocities
are reasonable – they provide a possibility for a first velocity pick. However, QC of the
picked velocities afterwards is essential. The following interactive velocity analysis script
provides three QC tools:
• a graph of the velocity profile,
• a plot of the NMO-corrected CMP, and
• a plot of the stack trace repeated eight times.
The velocity profile should be smooth and it should show generally increasing
velocities with depth. While NMO-velocity inversions (decreases) are possible, those
picks should be double-checked. The velocity profile should also show a pick near zero
time and a pick near the last time. This last requirement ensures that when sunmo

7-12
 Model 4: Sort, Velocity Analysis

interpolates velocity functions between CMPs, the interpolation is guided throughout the
time section.
The corrected CMP should exhibit flat primary reflections. After NMO is applied, an
upward curved primary reflection means the picked velocity is too slow (overcorrected)
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

and a downward curved primary reflection means the picked velocity is to fast (under-
corrected).
A fourth QC tool is provided when you decide to re-pick a CMP: the previous picks
are placed on the velan. This allows you to easily use or avoid previous picks.
Finally, how frequently along the seismic line must you do velocity analysis? It
depends on the complexity of the geology.

7.6.5 Computer Note 11: Interactive Scripts

Throughout this Primer, you will see that we have two methods to run an interactive
session. The two methods do not affect how data are processed. We prefer the second
method because we find it convenient to open a separate window for interactive
processing.
Method 1: Run a .sh script. For example,
iva.sh
When we use this method, the terminal window in which we execute this
command becomes the interactive dialog window.
Note: When using this method, never use the ampersand (&). That is, never use
the following command:
iva.sh &
As explained in Section 1.9.1, “&” puts the process in the background, something
you do not want to do when you are trying to be interactive!
Method 2: Run a .scr script. For example,
idecon.scr &
Here, the ampersand is optional, as explained in Section 1.9.
When we use this method, the .scr script opens an interactive dialog window and
starts a .sh script. Below is the iva.scr script:
1 #! /bin/sh
2 # File: iva.scr
3# Run this script to start script iva.sh
4
5 xterm -geom 8x20+10+545 -e iva.sh
6
This script has two active lines. Line 1 invokes the shell. Line 5 opens the
interactive window and starts the iva.sh script. In line 5, “80x20” specifies the size
of the dialog window: 80 characters wide by 20 characters high. Also, “10+545”
specifies the placement of the upper left corner of the dialog window: 10 pixels from
the left side of the viewing area and 545 pixels from the top of the viewing area.

7.6.6 Velocity Analysis: Scripts iva.scr and iva.sh

The following sections examine iva in detail. We would be pleased if you read every
sentence; however, you can use iva without doing so. We suggest the following as

7-13
 Model 4: Sort, Velocity Analysis

minimum reading: Section 7.6.6.2 (iva Displays) and Section 7.6.6.4 (iva.sh – User-
supplied Values).

7.6.6.1 iva.scr
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

You can do velocity analysis by creating a velan for each selected CMP along a line
(script oz14velan.sh, Section 7.6.2), then picking t-v pairs from the velan. After you have
t-v pairs for those selected CMPs, you can put them into a script that does NMO for the
line by interpolating between analyzed CMPs. However, if you do not like the flattened
CMPs or the later stack section, this style of velocity analysis is tedious to repeat. Script
iva.sh allows you to make picks interactively, view the flattened CMPs, and if necessary,
refine your picks.
As we explained in the previous section, we start interactive scripts by running a .scr
script. Here is iva.scr:
1 #! /bin/sh
2 # File: iva.scr
3 # Run this script to start script iva.sh
4
5 xterm -geom 80x20+10+545 -e iva.sh
6
Line 1 invokes the shell. Line 5 opens a dialog window of specified size and position.
Line 5 also starts iva.sh, the processing script, within that window.

7.6.6.2 iva Displays

Before looking at iva.sh, let’s look at the displays it creates. For the discussion in this
section, we limit processing to two CMPs: CMP 60 and CMP 80.
We start the interactive velocity analysis script by entering:
iva.scr &
The script prints the following information in the dialog window (as well as other
system messages that we do not show) and creates the plots shown in Figure 7.9.
*** INTERACTIVE VELOCITY ANALYSIS ***

Preparing CMP 1 of 2 for Picking

Location is CMP 60
Start CVS CMP = 55 End CVS CMP = 65

Use the semblance plot to pick (t,v) pairs.

Type "s" when the mouse pointer is where you want a pick.
Be sure your picks increase in time.
To control velocity interpolation, pick a first value
near zero time and a last value near the last time.
Type "q" in the semblance plot when you finish picking.
Following the directions above, we click on the border of the velan box to activate it,
then we make a series of picks in the velan image:
• Put the mouse pointer over a time near zero and a velocity near that of water
(because we know the top layer in the model is water (velocity = 1500 m/s), then
press letter “S.”

7-14
 Model 4: Sort, Velocity Analysis

• Put the mouse pointer over the first major event, at approximately 0.375 seconds
and velocity approximately 1500 m/s, then press letter “S.”
• Put the mouse pointer over the second major event, at approximately 0.7 seconds
and velocity approximately 2700 m/s, then press letter “S.”
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

• Put the mouse pointer over the third major event, at approximately 1.09 seconds
and velocity approximately 2900 m/s, then press letter “S.”
• You picked all the major events. Now make a final pick near the end time:
Following the t-v trend of the last two picks, put the mouse pointer near the last
time, 2 seconds, then press letter “S.”
• We are finished picking t-v pairs, so we press letter “Q.”

Figure 7.9: Left: The velan. Middle: The CVS plot. Right: A plot of the input CMP.

7-15
 Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

7-16
Model 4: Sort, Velocity Analysis

Figure 7.10: After picking. Left: The t-v graph. Middle-left: The CVS plot. Middle-right: The CMP after NMO. Right:
The stack trace repeated eight times.
 Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

7-17
Figure 7.11: The display for re-picking. Left: The color velan with the previous picks superimposed. Middle-left: The CVS plot.
Model 4: Sort, Velocity Analysis

Middle-right: The CMP after NMO. Right: A plot of the input CMP.
 Model 4: Sort, Velocity Analysis

The other plots are not just for show. At any time during picking, you can click the
middle mouse button in any window to get information. For example, by clicking near
the apex of the third reflector in the CMP window (Figure 7.12 below), the time, 1.06596
seconds tells you where to expect the third reflector to show maximum semblance in the
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

velan. You can then use the middle mouse button to click in the velan window to find the
semblance maximum that corresponds to 1.06596 seconds. (Are the numbers really
accurate to five decimal places? No.)

Figure 7.12: The upper left corner of a plot shows y,x information when you click the
middle mouse button in a window. Left: In the CMP window, y,x corresponds to
time,offset. Right: In the velan window, y,x corresponds to time,velocity.
Note: Whenever an suximage window is active, you can scroll through color palettes
by repeatedly pressing “r,” “R,” “h,” or “H.”. The following are lines from the ximage
selfdoc. (RGB stands for Red, Green, Blue; HSV stands for Hue, Saturation, Value.)
X Functionality:
Button 1 Zoom with rubberband box
Button 2 Show mouse (x1,x2) coordinates while pressed
q or Q key Quit
s key Save current mouse (x1,x2) location to file
p or P key Plot current window with pswigb (only from disk files)

... change colormap interactively

r install next RGB - colormap
R install previous RGB - colormap
h install next HSV - colormap
H install previous HSV - colormap
H install previous HSV - colormap

7-18
 Model 4: Sort, Velocity Analysis

(Move mouse cursor out and back into window for r,R,h,H to take effect)
Once “Q” is pressed when the velan window is active, the plots of Figure 7.10 appear.
The t-v graph looks good because (1) it starts at the top of the graph and ends near the
bottom, and (2) it is smooth, and (3) it generally increases with depth.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

The middle-right plot, the CMP after NMO, is also pleasing because it shows flat
reflectors.
The right plot, the repeat stack, shows strong amplitude peaks for the reflectors.
The dialog window shows the following:
t-v PICKS CMP 60
----------------------
0.0211082 1526.7
0.37467 1526.7
0.69657 2685
1.08707 2952.3
1.98945 3115.65

Use the velocity profile (left),

the NMO-corrected gather (middle-right),
and the repeated stack trace (right)
to decide whether to re-pick the CMP.

Picks OK? (y/n)

The dialog window list of t-v picks confirms that we have a pick near zero time
(0.0211082 seconds) and a pick near the end time (1.98945 seconds).
The picks are good, but we want to see the re-picking plots. So, we press “N” then the
Enter or Return key. The dialog window shows the following:
Picks OK? (y/n)
n

Repick CMP 60. Overlay previous picks.

Preparing CMP 1 of 2 for Picking

Location is CMP 60
Start CVS CMP = 55 End CVS CMP = 65

Use the semblance plot to pick (t,v) pairs.

Type "s" when the mouse pointer is where you want a pick.
Be sure your picks increase in time.
To control velocity interpolation, pick a first value
near zero time and a last value near the last time.
Type "q" in the semblance plot when you finish picking.
Figure 7.11 shows that the velan re-appears with an overlay of the picks. Also, the
original CMP appears on the right, covering the repeat stack plot.
We repeat our picks, guided by the white line. When we are finished, we press “Q.”
Plots similar to Figure 10 appear and the dialog window shows:
t-v PICKS CMP 60
----------------------
0.0158311 1497
0.369393 1526.7
0.69657 2670.15
1.08707 2952.3
1.98945 3130.5

Use the velocity profile (left),

the NMO-corrected gather (middle-right),

7-19
 Model 4: Sort, Velocity Analysis

and the repeated stack trace (right)

to decide whether to re-pick the CMP.

Picks OK? (y/n)

We are finished picking CMP 60, so we can enter “Y” or merely press Enter or
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Return. (You will see this below in script iva.sh.) We press Return. The dialog window
shows:
-- CLOSING CMP 60 WINDOWS --

Preparing CMP 2 of 2 for Picking

Location is CMP 80
Start CVS CMP = 75 End CVS CMP = 85

Use the semblance plot to pick (t,v) pairs.

Type "s" when the mouse pointer is where you want a pick.
Be sure your picks increase in time.
To control velocity interpolation, pick a first value
near zero time and a last value near the last time.
Type "q" in the semblance plot when you finish picking.
After a CMP is picked and before the plots for the next CMP are created, all plot
windows are closed.
Notice that a new series of CMPs is used to create the CVS plot for CMP 80.
After we “pick” CMP 80, and approve the picks (enter “Y” or merely press Enter or
Return when asked “Picks OK? (y/n)”), the dialog window shows:
-- CLOSING CMP 80 WINDOWS --

The output file of t-v pairs is vpick.txt

press return key to continue

After pressing Return or Enter, the dialog window closes.
Note: If, after a second try, we did not like our picks, we would repeat our “N”
answer to “Picks OK? (y/n)” and we would see plots as shown in Figure 11. This time the
velan would have an overlay of our second set of picks.
Note: The output file of t-v picks is called “vpick.txt.” We recommend that you
immediately change the name of this file so it will not be overwritten if you re-run
iva.scr. Alternatively, you can change the name of the “outpicks” file in script iva.sh.
Here is file vpicks.txt:
cdp=60,80
#=1,2
tnmo=0.0158311,0.369393,0.69657,1.08707,1.98945
vnmo=1497,1526.7,2670.15,2952.3,3130.5
tnmo=0.0158311,0.353562,0.69657,1.05541,1.98945
vnmo=1497,1541.55,2655.3,2774.1,2878.05
After minor editing, this file can be used by sunmo.

7.6.6.3 iva.sh — Introduction

Script iva.sh has the following major sections:
• Lines 8-47 are user-supplied values.
• Lines 48-83 document semblance (velan) and CVS computations. This section
includes computing the last semblance velocity (parameter lvs, lines 57-59).

7-20
 Model 4: Sort, Velocity Analysis

• Lines 84-118 document files created by iva.sh.

• Lines 119-159 do some one-time processing. Old temporary files are removed in
line 127. Old files can be left on the system when we crash the script. Remember,
you can terminate the script at any time by pressing “Control-C.” If you do not
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

plan to re-run the script right away, you can remove the temporary files by
entering a command similar to line 127:
rm panel.* picks.* par.* tmp*
Line 158 sets variable repick to “false” since, at this point, the user has not yet
made a first pick.
• Lines 160-440 are the core processing and plotting.
• Lines 441-453 create the velocity picks output file (outpicks).
• Lines 454-461 write a message to the screen telling us the name of the output file
of picks, pause the script, remove all temporary files, and close the script. The
pause of line 459 also gives us the chance to examine the temporary files before
they are removed.
1 #! /bin/sh
2 # File: iva.sh
3 # Run script iva.scr to start this script
4
5 # Set messages on
6 ##set -x
7
8 #================================================
9 # USER AREA -- SUPPLY VALUES
10 #------------------------------------------------
11 # CMPs for analysis
12
13 cmp1=60 cmp2=80 cmp3=100
14 cmp4=120 cmp5=140 cmp6=160
15 cmp7=180 cmp8=200 cmp9=220
16 cmp10=240 cmp11=260 cmp12=280
17 cmp13=300 cmp14=320 cmp15=340
18 cmp16=360 cmp17=380 cmp18=400
19
20 numCMPs=18
21
22 #------------------------------------------------
23 # File names
24
25 indata=cmp4.su # SU format
26 outpicks=vpick.txt # ASCII file
27
28 #------------------------------------------------
29 # display choices
30
31 myperc=98 # perc value for plot
32 plottype=0 # 0 = wiggle plot, 1 = image plot
33
34 #------------------------------------------------
35 # Processing variables
36
37 # Semblance variables
38 nvs=100 # number of velocities
39 dvs=27 # velocity intervals
40 fvs=1200 # first velocity
41

7-21
 Model 4: Sort, Velocity Analysis

42 # CVS variables
43 fc=1200 # first CVS velocity
44 lc=3900 # last CVS velocity
45 nc=10 # number of CVS velocities (panels)
46 XX=11 # ODD number of CMPs to stack into central CVS
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

47
48 #================================================
49
50 # HOW SEMBLANCE (VELAN) VELOCITIES ARE COMPUTED
51
52 # Last Vel = fvs + (( nvs-1 ) * dvs ) = lvs
53 # 5000 = 500 + (( 99-1 ) * 45 )
54 # 3900 = 1200 + (( 100-1 ) * 27 )
55
56 # Compute last semblance (velan) velocity
57 lvs=`bc -l << -END
58 $fvs + (( $nvs - 1 ) * $dvs )
59 END`
60
61 #------------------------------------------------
62
63 # HOW CVS VELOCITIES ARE COMPUTED
64
65 # dc = CVS velocity increment
66 # dc = ( last CVS vel - first CVS vel ) / ( # CVS - 1 )
67 # m = CVS plot trace spacing (m = d2, vel units)
68 # m = ( last CVS vel - first CVS vel ) / ( ( # CVS - 1 ) * XX )
69
70 # j=1
71 # while [ j le nc ]
72 # do
73 # vel = fc + { [( lc - fc ) / ( nc-1 )] * ( j-1) }
74 # j = j + 1
75 # done
76 # EXAMPLE:
77 # vel = 1200+ ( (( 3900 - 1200 ) / ( 10-1 )) * ( 1-1) )
78 # vel = 1200+ ( (( 3900 - 1200 ) / ( 10-1 )) * ( 2-1) )
79 # .
80 # .
81 # .
82 # vel = 1200 + ( (( 3900 - 1200 ) / ( 10-1 )) * (11-1) )
83
84 #================================================
85
86 # FILE DESCRIPTIONS
87
88 # tmp0 = binary temp file for input CVS gathers
89 # tmp1 = binary temp file for output CVS traces
90 # tmp2 = ASCII temp file for managing picks
91 # tmp3 = binary temp file for stacked traces
92 # tmp4 = ASCII temp file for "wc" result (velan)
93 # tmp5 = ASCII temp file for stripping file name from tmp4 (velan)
94 # tmp6 = ASCII temp file to avoid screen display of "zap"
95 # tmp7 = ASCII temp file for picks
96 # tmp8 = binary temp file for NMO (flattened) section
97 # panel.$picknow = current CMP windowed from line of CMPs
98 # picks.$picknow = current CMP picks arranged as "t1 v1"
99 # "t2 v2"
100 # etc.
101 # par.# (# is a sequential index number; 1, 2, etc.)
102 # = current CMP picks arranged as
103 # "tnmo=t1,t2,t3,...
104 # "vnmo=v1,v2,v3,...

7-22
 Model 4: Sort, Velocity Analysis

105 # par.uni.# (# is a sequential index number; 1, 2, etc.)

106 # = current CMP picks arranged as
107 # "xin=t1,t2,t3,...
108 # "yin=v1,v2,v3,...
109 # for input to xgraph to display velocity profile
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

110 # par.cmp = file of CMP number and sequential index number;

111 # for example: "40 1"
112 # "60 2"
113 # etc.
114 # par.0 = file "par.cmp" re-arranged as
115 # "cdp=#,#,#,etc." NOTE: # in this line is picked CMP
116 # "#=1,2,3,etc." NOTE: # in this line is "#"
117 # outpicks = concatenation of par.0 and all par.# files.
118
119 #================================================
120
121 echo " "
122 echo " *** INTERACTIVE VELOCITY ANALYSIS ***"
123 echo " "
124
125 #------------------------------------------------
126 # Remove old files. Open new files
127 rm -f panel.* picks.* par.* tmp*
128
129 > $outpicks # Write empty file for final picks
130 > par.cmp # Write empty file for recording CMP values
131
132 #------------------------------------------------
133 # Get ns, dt, first time from seismic file
134 nt=`sugethw ns < $indata | sed 1q | sed 's/.*ns=//'`
135 dt=`sugethw dt < $indata | sed 1q | sed 's/.*dt=//'`
136 ft=`sugethw delrt < $indata | sed 1q | sed 's/.*delrt=//'`
137
138 # Convert dt from header value in microseconds
139 # to seconds for velocity profile plot
140 dt=`bc -l << -END
141 scale=6
142 $dt / 1000000
143 END`
144
145 # XGRAPH: If "delrt", use it; else use zero
146 if [ $ft -ne 0 ] ; then
147 tstart=`bc -l << -END
148 scale=6
149 $ft / 1000
150 END`
151 else
152 tstart=0.0
153 fi
154
155 #------------------------------------------------
156
157 # Initialize "repick" -- for plotting previous picks on velan
158 repick=1 # 1=false, 0=true
159
160 #------------------------------------------------
161 # BEGIN IVA LOOP
162 #------------------------------------------------
163
164 i=1
165 while [ $i -le $numCMPs ]
166 do
167

7-23
 Model 4: Sort, Velocity Analysis

168 # set variable $picknow to current CMP

169 eval picknow=\$cmp$i
170
171 if [ $repick -eq 1 ] ; then
172 echo " "
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

173 echo "Preparing CMP $i of $numCMPs for Picking "

174 echo "Location is CMP $picknow "
175 fi
176
177 #------------------------------------------------
178 # Plot CMP (right)
179 #------------------------------------------------
180
181 suwind < $indata \
182 key=cdp min=$picknow max=$picknow > panel.$picknow
183 if [ $repick -eq 1 ] ; then
184 if [ $plottype -eq 0 ] ; then
185 suxwigb < panel.$picknow xbox=634 ybox=10 wbox=300 hbox=450 \
186 title="CMP gather $picknow" \
187 label1=" Time (s)" label2="Offset (m)" key=offset \
188 perc=$myperc verbose=0 &
189 else
190 suximage < panel.$picknow xbox=634 ybox=10 wbox=300 hbox=450 \
191 title="CMP gather $picknow" \
192 label1=" Time (s)" \
193 perc=$myperc verbose=0 &
194 fi
195 else
196 if [ $plottype -eq 0 ] ; then
197 suxwigb < panel.$picknow xbox=946 ybox=10 wbox=300 hbox=450 \
198 title="CMP gather $picknow" \
199 label1=" Time (s)" label2="Offset (m)" key=offset \
200 perc=$myperc verbose=0 &
201 else
202 suximage < panel.$picknow xbox=946 ybox=10 wbox=300 hbox=450 \
203 title="CMP gather $picknow" \
204 label1=" Time (s)" \
205 perc=$myperc verbose=0 &
206 fi
207 fi
208
209 #------------------------------------------------
210 # Constant Velocity Stacks (CVS) (middle-left)
211 # Make CVS plot for first pick effort.
212 # If re-picking t-v values, do not make this plot.
213 #------------------------------------------------
214
215 # repick: 1=false, 0=true
216 if [ $repick -eq 1 ] ; then
217
218 # number of CMPs - 1; for windowing
219 X=`expr $XX - 1`
220
221 # Window CMPs around central CMP (+/- X/2). Write to tmp0
222 k1=`expr $picknow - $X / 2` # Window from CMP to CMP - X/2
223 k2=`expr $picknow + $X / 2` # Window from CMP to CMP + X/2
224 suwind < $indata key=cdp min=$k1 max=$k2 > tmp0
225
226 # Calculate CVS velocity increment
227 # dc = ( last CVS vel - first CVS vel ) / ( # CVS - 1 )
228 dc=`bc -l << -END
229 ( $lc - $fc ) / ( $nc - 1 )
230 END`

7-24
 Model 4: Sort, Velocity Analysis

231
232 # Calculate trace spacing for CVS plot (m = d2, vel units)
233 # m = ( last CVS vel - first CVS vel ) / ( ( # CVS - 1 ) * XX )
234 m=`bc -l << -END
235 ( $lc - $fc ) / ( ( $nc - 1 ) * $XX )
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

236 END`
237
238 # CVS velocity loop
239 j=1
240 while [ $j -le $nc ]
241 do
242
243 vel=`bc -l << -END
244 $fc + $dc * ( $j - 1 )
245 END`
246
247 # uncomment to print CVS velocities to screen
248 ## echo " vel = $vel"
249
250 sunmo < tmp0 vnmo=$vel verbose=0 |
251 sustack >> tmp1
252
253 j=`expr $j + 1`
254 done
255
256 # Compute lowest velocity for annotating CVS plot
257 # loV = first CVS velocity - ( ( CMP range - 1 ) / 2 ) * vel inc
258 loV=`bc -l << -END
259 $fc - ( $X / 2) * $m
260 END`
261
262 suximage < tmp1 xbox=322 ybox=10 wbox=300 hbox=450 \
263 title="CMP $picknow Constant Velocity Stacks" \
264 label1=" Time (s)" label2="Velocity (m/s)" \
265 f2=$loV d2=$m verbose=0 \
266 perc=$myperc n2tic=5 cmap=rgb0 &
267
268 fi
269
270 #------------------------------------------------
271 # Picking instructions
272 #------------------------------------------------
273
274 echo " "
275 echo "Preparing CMP $i of $numCMPs for Picking "
276 echo "Location is CMP $picknow "
277 echo " Start CVS CMP = $k1 End CVS CMP = $k2"
278 echo " "
279 echo " Use the semblance plot to pick (t,v) pairs."
280 echo " Type \"s\" when the mouse pointer is where you want a pick."
281 echo " Be sure your picks increase in time."
282 echo " To control velocity interpolation, pick a first value"
283 echo " near zero time and a last value near the last time."
284 echo " Type \"q\" in the semblance plot when you finish picking."
285
286 #------------------------------------------------
287 # Plot semblance (velan) (left)
288 #------------------------------------------------
289
290 # repick: 1=false, 0=true
291 if [ $repick -eq 0 ] ; then
292
293 #--- --- --- --- --- --- --- --- --- ---

7-25
 Model 4: Sort, Velocity Analysis

294 # Get the number of picks (number of lines) in tmp7 |

295 # Remove blank spaces preceding the line count.
296 # Remove file name that was returned from "wc".
297 # Store line count in "npair" to guide line on velan.
298
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

299 wc -l tmp7 | sed 's/^ *\(.*\)/\1/' > tmp4

300 sed 's/tmp7//' < tmp4 > tmp5
301 npair=`sort < tmp5`
302 #--- --- --- --- --- --- --- --- --- ---
303
304 suvelan < panel.$picknow nv=$nvs dv=$dvs fv=$fvs |
305 suximage xbox=10 ybox=10 wbox=300 hbox=450 perc=99 \
306 units="semblance" f2=$fvs d2=$dvs n2tic=5 \
307 title="Semblance Plot CMP $picknow" cmap=hsv2 \
308 label1=" Time (s)" label2="Velocity (m/s)" \
309 legend=1 units=Semblance verbose=0 gridcolor=black \
310 grid1=solid grid2=solid mpicks=picks.$picknow \
311 curve=tmp7 npair=$npair curvecolor=white
312
313 else
314
315 suvelan < panel.$picknow nv=$nvs dv=$dvs fv=$fvs |
316 suximage xbox=10 ybox=10 wbox=300 hbox=450 perc=99 \
317 units="semblance" f2=$fvs d2=$dvs n2tic=5 \
318 title="Semblance Plot CMP $picknow" cmap=hsv2 \
319 label1=" Time (s)" label2="Velocity (m/s)" \
320 legend=1 units=Semblance verbose=0 gridcolor=black \
321 grid1=solid grid2=solid mpicks=picks.$picknow
322
323 fi
324
325 #------------------------------------------------
326 # End first set of plots
327 #================================================
328
329 #------------------------------------------------
330 # Manage picks (1): Prepare picks for sunmo
331 #------------------------------------------------
332
333 sort < picks.$picknow -n |
334 mkparfile string1=tnmo string2=vnmo > par.$i
335 echo "cdp=$picknow" >> tmp2
336 cat par.$i >> tmp2
337
338 #================================================
339 # Begin second set of plots
340 #------------------------------------------------
341
342 #------------------------------------------------
343 # Flattened seismic data (NMO) plot (middle-right)
344 #------------------------------------------------
345
346 sunmo < panel.$picknow par=tmp2 verbose=0 > tmp8
347 if [ $plottype -eq 0 ] ; then
348 suxwigb < tmp8 xbox=634 ybox=10 wbox=300 hbox=450 \
349 title="CMP $picknow after NMO" \
350 label1=" Time (s)" label2="Offset (m)" \
351 verbose=0 perc=$myperc key=offset &
352 else
353 suximage < tmp8 xbox=634 ybox=10 wbox=300 hbox=450 \
354 title="CMP $picknow after NMO" \
355 label1=" Time (s)" \
356 verbose=0 perc=$myperc &

7-26
 Model 4: Sort, Velocity Analysis

357 fi
358
359 #------------------------------------------------
360 # Stack window (right)
361 #------------------------------------------------
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

362
363 j=1
364 while [ $j -le 8 ]
365 do
366
367 # Append stack trace into tmp3 multiple times
368 sustack < tmp8 >> tmp3
369
370 j=`expr $j + 1`
371 done
372
373 suxwigb < tmp3 xbox=946 ybox=10 wbox=200 hbox=450 \
374 title="CMP $picknow repeat stack trace" \
375 label1=" Time (s)" d2num=50 key=cdp \
376 verbose=0 perc=$myperc &
377
378 #------------------------------------------------
379 # Manage picks (2): Prepare picks for vel profile
380 #------------------------------------------------
381
382 sed < par.$i '
383 s/tnmo/xin/
384 s/vnmo/yin/
385 ' > par.uni.$i
386
387 #------------------------------------------------
388 # Velocity profile (left)
389 #------------------------------------------------
390
391 unisam nout=$nt fxout=$tstart dxout=$dt \
392 par=par.uni.$i method=mono |
393 xgraph n=$nt nplot=1 d1=$dt f1=$tstart x2beg=$fvs x2end=$lvs \
394 label1=" Time (s)" label2="Velocity (m/s)" \
395 title="CMP $picknow Stacking Velocity Function" \
396 -geometry 300x450+10+10 -bg white style=seismic \
397 grid1=solid grid2=solid linecolor=2 marksize=1 mark=0 \
398 titleColor=black axesColor=blue &
399
400 #------------------------------------------------
401 # Dialogue with user: re-pick ?
402 #------------------------------------------------
403
404 echo " "
405 echo " t-v PICKS CMP $picknow"
406 echo "----------------------"
407 cat picks.$picknow
408 echo " "
409 echo " Use the velocity profile (left),"
410 echo " the NMO-corrected gather (middle-right),"
411 echo " and the repeated stack trace (right)"
412 echo " to decide whether to re-pick the CMP."
413 echo " "
414 echo "Picks OK? (y/n) " > /dev/tty
415 read response
416
417 rm tmp*
418
419 # "n" means re-loop. Otherwise, continue to next CMP.

7-27
 Model 4: Sort, Velocity Analysis

420 case $response in

421 n*)
422 i=$i
423 echo " "
424 echo "Repick CMP $picknow. Overlay previous picks."
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

425 repick=0
426 cp picks.$picknow tmp7
427 ;;
428 *)
429 echo "$picknow $i" >> par.cmp
430 i=`expr $i + 1`
431 repick=1
432 echo "-- CLOSING CMP $picknow WINDOWS --"
433 zap xwigb > tmp6
434 zap ximage > tmp6
435 zap xgraph > tmp6
436 ;;
437 esac
438
439 done
440
441 #------------------------------------------------
442 # Create velocity output file
443 #------------------------------------------------
444
445 mkparfile < par.cmp string1=cdp string2=# > par.0
446
447 i=0
448 while [ $i -le $numCMPs ]
449 do
450 sed < par.$i 's/$/ \\/g' >> $outpicks
451 i=`expr $i + 1`
452 done
453
454 #------------------------------------------------
455 # Remove files and exit
456 #------------------------------------------------
457 echo " "
458 echo " The output file of t-v pairs is "$outpicks
459 pause
460 rm -f panel.* picks.* par.* tmp*
461 exit
462

7.6.6.4 iva.sh — User-supplied Values

Line 20 has the number of CMPs to process, the number of times to loop through
velocity analysis. Line 20 is tied to the velocity analysis loop with the following lines:
20 numCMPs=18
164 i=1
165 while [ $i -le $numCMPs ]
166 do
430 i=`expr $i + 1`
439 done
Line 164 initializes counter variable “i”; line 430 increments counter “i” when we finish
picking a CMP and want to move to the next CMP or quit the script. Line 165 ensures
that only the number of CMPs in line 20 are processed.
Note: In line 20, you can use any value up to the total number of CMPs you list. We
sometimes use a value of “2” or “3” when we are testing the script or previewing the
data.

7-28
 Model 4: Sort, Velocity Analysis

Lines 13-18 list the CMPs. Here, the CMPs evenly increment by 20, but any values
are acceptable. The CMPs are listed on six lines, but they can be listed on three lines, two
lines, or even one per line to use 18 lines. The number per line is not important. However,
the style of these lines is very important. When the eval function operates, line 169, it
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

sets the value of picknow, the current CMP.

169 eval picknow=\$cmp$i
The first time line 169 is used, the value of picknow becomes the value of parameter
cmp1. The script searches its memory and finds that cmp1=60, so another part of the
script (lines 181-182) windows CMP 60 from the input file. The second time line 156 is
used, the value of picknow becomes the value of parameter cmp2. The script searches its
memory and finds that cmp2=80, so lines 181-182 extract CMP 80 from the input file.
Etc. Because line 169 uses “cmp,” lines 13-18 must also use “cmp.”
Note: The Bourne shell does not have “tables.” To get around this restriction, lines
13-18 list the table elements and line 20 has the number of table elements. Line 169
concatenates the $cmp variable with the value of the $i variable. Line 169 effectively
selects the appropriate array element. This pseudo-table structure allows us to specify
unequally spaced CMPs. We thank Brian Weeden for helping us understand eval.
Line 26 is the place for the name of the output t-v pick file. We suggest that if you do
not rename this file here with every run, that you rename the output file after each run.
Before you set the perc value in line 31, we suggest you determine a perc value by
using suxwigb or suximage externally.
Note: You might want to have a separate perc value for the occurrences of suximage
that follow suvelan (lines 304-321). For example, after line 31, add the following line:
percvelan=98 # perc value for velan ximage
Then, on lines 305 and 316, change perc=99 to perc=$percvelan. Using
percvelan, you can change the “amplitude” of the velan plots separately from the other
plots. This is a powerful tool for revealing details in real data.
Line 32 lets you set the CMP plot and the flattened CMP plot as wiggle trace plots or
image plots. This option is here because we find that “too many” traces do not display
well in suxwigb. (“Too many” is qualitative!)
Lines 37-46 let you set the semblance and CVS variables. To understand the
semblance variables, use the documentation on lines 52-54, study lines 304 and 315, and
read the suvelan selfdoc. To understand the CVS variables, use the documentation on
lines 63-82 and study lines 218-268.
Note: You can uncomment line 248 to print the CVS velocities to the screen (see
Section 7.6.3).

7.6.6.5 iva.sh — File Documentation

Lines 84-118 describe the files created by iva.sh. The small files are ASCII and the
large ones contain seismic data. During processing, you can investigate the contents of
the ASCII files by using cat and you can investigate the contents of the seismic files by
using suxwigb or suximage.

7-29
 Model 4: Sort, Velocity Analysis

7.6.6.6 iva.sh — One-time Processing

Line 127 removes old temporary files if a previous run that crashed did not permit
execution of line 460.
127 rm -f panel.* picks.* par.* tmp*
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

460 rm -f panel.* picks.* par.* tmp*

Line 129 opens the final output file. We open the file here to make sure the command
works instead of waiting until all processing is finished.
129 > $outpicks # Write empty file for final picks
Line 134 extracts the value of key ns from the first trace of the input file; line 135
extracts the value of key dt from the first trace of the input file. We use these values in
lines 391 and 393 to create the velocity profile plot.
134 nt=`sugethw ns < $indata | sed 1q | sed 's/.*ns=//'`
135 dt=`sugethw dt < $indata | sed 1q | sed 's/.*dt=//'`
Every trace has the same value for keys ns and dt (see the surange output of cmp4.su,
Section 7.2). The syntax of lines 134 and 135 can be used to extract the value of any key
from the first trace of the input seismic data.
Line 141, scale=6, sets the number of significant figures (6) of the computed dt
value, lines 140-143.
140 dt=`bc -l << -END
141 scale=6
142 $dt / 1000000
143 END`
Line 136 extracts the value of key delrt: delay recording time (milliseconds). As with
ns and dt, this is used for the velocity profile. If the input file was windowed in time, the
non-zero value of delrt is converted to seconds (lines 147-150) and used as parameter
tstart to label the beginning time of the velocity profile (lines 391-398). If delrt is not
used or contains zero, tstart is directly set to zero.
Because we check the value of delrt, and, if it is non-zero, we use it to set the start
time of the velocity profile, we can do velocity analysis on a time-windowed data set.
However, as was pointed out in Section 7.6.4, “The velocity profile should also show a
pick near zero time and a pick near the last time.” If a time-windowed file is used, after
picking, the pick file must be edited to include picks near zero time and near the last
times.

7.6.6.7 iva.sh — IVA Loop

This is a big section, so we examine it in blocks.
Lines 181-182 window a CMP from the input file.
If this is the first time we see the CMP; that is, this is not a re-pick (repick=1), the
upper left corner of the CMP plot is placed at x=634, y=10 (line 185 or 190). If we are
viewing the CMP after we picked t-v pairs, the CMP plot is placed farther to the right so
it does not cover the flattened (NMO) plot. (The flattened CMP plot is created and plotted
in lines 346-357.)
To make a CVS panel, a single velocity flattens an odd number of CMPs, then each
CMP is stacked, then all the stack traces are displayed side-by-side. Line 46 (variable XX)
specifies that eleven CMPs are to be used for each panel. Line 45 (variable nc) specifies

7-30
 Model 4: Sort, Velocity Analysis

the number of single velocities to apply to those eleven CMPs; in other words, the
number of panels.
45 nc=10 # number of CVS velocities (panels)
46 XX=11 # ODD number of CMPs to stack into central CVS
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

If the first CMP is 60, CMP 60 is the panel’s center CMP. Lines 218-223 compute the
minimum and maximum CMP values of this series of eleven CMPs. Line 224 windows
the eleven CMPs from the input file and puts them in file tmp0. In lines 239-254, as each
of the ten velocities is computed, NMO is applied to tmp0 and the CMPs are stacked.
Line 251 uses the append command to concatenate successive panels into file tmp1.
251 sustack >> tmp1
File, tmp1 is plotted by suximage, lines 262-266. The CVS horizontal axis is velocity.
Lines 234-236 compute velocity trace spacing (m) and lines 258-260 compute the first
(lowest) annotation velocity (loV)for the CVS suximage plot (line 265).
Lines 274-284 print the picking instructions to the screen. They also print information
about the CMP that is about to be displayed (variables numCMPs, picknow, k1, and k2 in
lines 275-277). Notice that lines 173-174 previously printed this information. We repeat
the information because system messages printed when sunmo runs scrolls earlier text
out of the viewing area.
Lines 315-321 create and plot the velan if this is the first time the velan is created for
the current CMP. Lines 299-311 are used if the velan is being displayed for re-picking. In
both cases, user-supplied values for nvs, dvs, and fvs are used (lines 304, 306, 315, 317).
For re-picking, we have to tell suximage, in line 311, the name of the file of the previous
picks (parameter curve) and how many t-v picks are in the file (parameter npair).
311 curve=tmp7 npair=$npair curvecolor=white
Why does file tmp7 hold the picks? Here we explain how picks are made and how
they get into tmp7. Notice that suximage, suxwigb, and xgraph always end with an
ampersand (&) except the two times suximage is used for picking: lines 305-311 and
lines 316-321. In these two uses of suximage, parameter mpicks is used and has the value
picks.$picknow:
310 grid1=solid grid2=solid mpicks=picks.$picknow \
321 grid1=solid grid2=solid mpicks=picks.$picknow
Because these uses of suximage do not use the ampersand (&), script control goes to
ximage. When the velan plot is active and you press “Q,” control leaves ximage and
returns to the script (see the ximage selfdoc).
Whether we are picking a CMP the first time, the second time, or … the picks go in
the same file. However, every time we choose to re-pick (the re-pick yes/no case is lines
420-437), the old picks are copied to tmp7 by line 426:
426 cp picks.$picknow tmp7
So, tmp7 has the old picks to overlay on the velan while new picking is going on.
Now we return to our travel down the script.
Lines 299-301 put the number of t-v picks (the number of lines) of tmp7 into variable
npair. Unfortunately, command wc –l also outputs the file name. For example, wc –l with
the iva.sh script returns:
·····449 iva.sh

7-31
 Model 4: Sort, Velocity Analysis

Those dots show that the output line count (449) is preceded by five blank spaces. The
second part of line 299, the sed command, removes the preceding blank spaces (Lamb,
1990, page 91, number 11). We know that the name of the t-v pick file is tmp7; we use
that information on line 300 to remove the file name from tmp4. The sort on line 301 is
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

merely a device to copy the contents of tmp5 (the line count) to variable npair.
After picking (after “Q” is pressed), lines 333-336 create a parameter (par) file
formatted for input to sunmo. Command mkparfile is an SU function that has a selfdoc.
The following are the contents of tmp2 after picking CMP 60:
cdp=60
tnmo=0.0158311,0.369393,0.69657,1.08707,1.98945
vnmo=1497,1526.7,2670.15,2952.3,3130.5
File tmp2 is used by sunmo in line 346 to make the flattened CMP for a QC display.
The flattened CMP, file tmp8, is used by the next QC display, the repeat stack trace (lines
363-376). You might want fewer or more traces than the eight specified in line 364.
Lines 382-392 create a par file for the QC t-v graph. While sunmo requires tnmo-
vnmo pairs, xgraph requires xin-yin pairs (lines 383-384). SU program unisam creates a
smooth line from a series of point pairs (lines 391-392).
By the time we arrive at line 415, we are viewing the t-v graph, the CVS plot, the
flattened CMP, and the repeat stacks. Also, the dialog window shows the current t-v pairs
(line 407). Here, we have to choose whether to re-pick the CMP or accept our picks. No
matter which we choose, all files with the tmp prefix are removed at line 417 to clear the
files for the next round of displays.
If we choose to re-pick, the old picks are copied to tmp7 for overlay on the next velan
(line 426). If we choose to accept the current picks, the CMP counter increments (line
430), and all the plot windows are closed by lines 433-435. These lines direct (>) the
command zap to file tmp6 in order to minimize system messages to the screen.

7.6.6.8 iva.sh — Output the Picks

At the finish of processing, file par.0 is created (line 445). Then (lines 447-452), file
par.0 is concatenated with files par.1 (the file of tnmo-vnmo pairs for CMP 60), par.2 (the
file of tnmo-vnmo pairs for CMP 80), etc., into file outpick.
Line 445 uses SU command mkparfile to re-orient the columns of par.cmp to rows.
File par.cmp holds the CMP numbers and their processing order:
60 1
80 2
100 3
etc.
Below is an example output file after processing two CMPs (line20: numCMPs=2).
cdp=60,80 \
#=1,2 \
tnmo=0.0158311,0.369393,0.69657,1.08707,1.98945 \
vnmo=1497,1526.7,2670.15,2952.3,3130.5 \
tnmo=0.0158311,0.353562,0.69657,1.05541,1.98945 \
vnmo=1497,1541.55,2655.3,2774.1,2878.05 \
Line 450 puts a blank space and a continuation mark, the backslash (\), at the end of each
line (Lamb, 1990, pages 91-92).

7-32
 Model 4: Sort, Velocity Analysis

7.6.6.9 iva.sh — Pause, Remove Temporary Files, Exit

Just before removing all temporary files and exiting, the name of the output file is
written to the screen. We suggest that you rename the file to prevent writing over the
picks with a later run.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

7.6.6.10 iva.sh — vpick.txt

The seismic data input to interactive velocity analysis is cmp4.su. The output file of t-
v picks is below, In Chapter 8, we call this file vpick4.txt.
1 cdp=60,80,100,120,140,160,180,200,220,240,260,280,300,320,340,360,380,400 \
2 #=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18 \
3 tnmo=0.0158311,0.390501,0.686016,1.06069,1.99472 \
4 vnmo=1511.85,1511.85,2640.45,2922.6,3249.3 \
5 tnmo=0.0158311,0.353562,0.701847,1.0343,1.99472 \
6 vnmo=1482.15,1511.85,2625.6,2774.1,2982 \
7 tnmo=0.0158311,0.364116,0.765172,0.875989,1.08707,1.98945 \
8 vnmo=1511.85,1526.7,2595.9,2714.7,2625.6,2699.85 \
9 tnmo=0.0211082,0.437995,0.844327,0.949868,1.16623,1.98945 \
10 vnmo=1497,1526.7,2551.35,2432.55,2685,2685 \
11 tnmo=0.0105541,0.569921,1.00792,1.30343,1.99472 \
12 vnmo=1497,1541.55,2595.9,2848.35,3011.7 \
13 tnmo=0.0158311,0.686016,1.12401,1.44591,1.98945 \
14 vnmo=1482.15,1541.55,2729.55,3160.2,3293.85 \
15 tnmo=0.0158311,0.78628,1.19789,1.5409,1.99472 \
16 vnmo=1482.15,1511.85,3011.7,3694.8,3798.75 \
17 tnmo=0.0105541,0.765172,1.1029,1.45119,1.98945 \
18 vnmo=1497,1511.85,2937.45,3665.1,3858.15 \
19 tnmo=0.0158311,0.627968,0.907652,1.25594,1.99472 \
20 vnmo=1497,1556.4,2595.9,3056.25,3353.25 \
21 tnmo=0.0158311,0.469657,0.686016,1.07124,1.98417 \
22 vnmo=1482.15,1556.4,2521.65,2818.65,3026.55 \
23 tnmo=0.0158311,0.3219,0.501319,0.923483,1.98417 \
24 vnmo=1482.15,1541.55,2462.25,2670.15,2774.1 \
25 tnmo=0.0158311,0.226913,0.395778,0.870712,1.98417 \
26 vnmo=1482.15,1511.85,2521.65,2699.85,2848.35 \
27 tnmo=0.0105541,0.23219,0.390501,0.918206,1.99472 \
28 vnmo=1482.15,1526.7,2566.2,2878.05,3234.45 \
29 tnmo=0.0158311,0.258575,0.46438,1.99472 \
30 vnmo=1497,1526.7,2699.85,3368.1 \
31 tnmo=0.0158311,0.311346,0.538259,0.612137,1.98945 \
32 vnmo=1482.15,1526.7,2759.25,2952.3,3442.35 \
33 tnmo=0.0211082,0.353562,0.622691,0.744063,1.99472 \
34 vnmo=1467.3,1511.85,2774.1,3056.25,3457.2 \
35 tnmo=0.0211082,0.369393,0.691293,0.812665,1.99472 \
36 vnmo=1482.15,1511.85,2878.05,2982,3279 \
37 tnmo=0.0211082,0.348285,0.728232,0.881266,1.20844,1.98945 \
38 vnmo=1482.15,1511.85,2833.5,3026.55,3145.35,3323.55 \

Note: Line 2 of this file must be removed before it can be used in a script. A comment
line is not acceptable within a script command (see Section 8.2).

7.7 X-axis Labels in suxwigb and suximage

Program suxwigb allows you to supply a key value for labeling the x-axis. (In SU, the
seismic x-axis is also called the slow dimension.) If you supply a key value; for example,
offset, suxwigb correctly annotates offset values. It even plots the traces with irregular
spacing if the key values do not occur at regular increments.
Program suximage accepts instructions for labeling the slow dimension (the seismic
x-axis). Program suximage also has default instructions for labeling the x-axis; that is, a
way to label the x-axis when the user does not supply an instruction.
Let’s look at a small part of the suximage code (suximage.c, 2005):

7-33
 Model 4: Sort, Velocity Analysis

153 if (!getparfloat("f2", &f2)) {

154 if (tr.f2) f2 = tr.f2;
155 else if (akey) f2 = vtof(stype,val1) ;
156 else if (tr.tracr) f2 = (float) tr.tracr;
157 else if (tr.tracl) f2 = (float) tr.tracl;
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

158 else if (seismic) f2 = 1.0;

159 else f2 = 0.0;
160 }
These lines are a hierarchy for creating the first value for the slow dimension.
• If the user supplied a value for parameter f2, use it.
• If the user supplied a value for key, use it.
• If trace header key tracr has a value, use it.
• If trace header key tracl has a value, use it.
• If this is seismic data, set the first value to “1.”
• If none of the preceding is true, set the first value to “0.”
In SU documentation, f2 is called “first trace location.” By extension, all the parameters
listed above only set the annotation value for the first displayed trace.
Now, let’s look at trace (slow dimension) increment. Below is the line of code in
suximage.c that refers to trace header key d2, “sample spacing between traces.”
145 if (!getparfloat("d2", &d2)) d2 = (tr.d2) ? tr.d2 : 1.0;
This line creates a hierarchy for setting d2.
If the user supplied a value for d2, use it.
If trace header key d2 has a value, use it.
If neither of the preceding is true, set d2 to “1.”
In conclusion:
• Program suxwigb reads each trace’s key value and properly increments irregularly
spaced values.
• Program suximage:
o Careful use of f2 and d2 can produce a result similar to the key parameter
of suxwigb.
o The default suximage increment is “1,”so trace labeling after the first
trace will be correct in only the simplest cases.

7-34
 Model 4: T-V Picks QC, NMO, Stack

8. Model 4: T-V Picks QC, NMO, Stack

8.1 Introduction
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Now that we have a file of time-velocity (t-v) picks at selected CMPs, we check the
quality of those picks. Then, we apply NMO to the data. After reviewing the results of
NMO, it is simple (a one-line command) to stack the gathers.
For the sake of presentation in Section 8.2.1, what was one line in vpick4.txt (line 1)
are now two lines in tvQC.sh (lines 13-14). Notice that THERE ARE NO SPACES AT
THE END OF LINE 13 AND NO SPACES AT THE BEGINNING OF LINE 14.
We removed line 2 from file vpick4.txt. That line would be between lines 14 and 15
below. We removed it because A COMMENT LINE WITHIN THE LINES OF AN SU
COMMAND WILL MAKE THE SCRIPT CRASH.
We also removed the continuation mark from the end of line 50.

8.2 Time-Velocity QC
At the end of interactive velocity analysis (Section 7.6.6.10), we have a file that
contains a row of cdp values. Also within that file, there is a pair of tnmo values and
vnmo values for each cdp value. Before we use these cdp-tnmo-vnmo values in normal
moveout correction (NMO), we want to check the quality of our picks.

8.2.1 Program tvnmoqc, mode = 1

Each time series (set of tnmo values) input to the Seismic Un*x NMO program,
sunmo, must increase in time. In other words, in a time series, each successive pick must
be at a later time than the previous pick. The following script, tvqc.sh, offers a fast way to
check that the tnmo values are acceptable to sunmo.
Script tvQC.sh uses program tvnmoqc; that program is presented in Appendix C. In
tvQC.sh, we set mode=1. In this mode, tvnmoqc checks the number of cdp values, it
checks that the number tnmo-vnmo pairs match the number of cdp values, and it checks
that values in a tnmo series always increase. The output of tvnmoqc is screen messages.
1 #! /bin/sh
2 # File: tvQC.sh
3 # Test tnmo-vnmo series
4 # Input: The same cdp-tnmo-vnmo file that
5 # would be input to sunmo
6 # Output: Screen messages
7 # Use: tvQC.sh
8
9 # Set debugging on
10 ##set -x
11
12 tvnmoqc mode=1 \
13 cdp=60,80,100,120,140,160,180,200,220,\
14 240,260,280,300,320,340,360,380,400 \
15 tnmo=0.0158311,0.390501,0.686016,1.06069,1.99472 \
16 vnmo=1511.85,1511.85,2640.45,2922.6,3249.3 \
17 tnmo=0.0158311,0.353562,0.701847,1.0343,1.99472 \
18 vnmo=1482.15,1511.85,2625.6,2774.1,2982 \
19 tnmo=0.0158311,0.364116,0.765172,0.875989,1.08707,1.98945 \
20 vnmo=1511.85,1526.7,2595.9,2714.7,2625.6,2699.85 \

8-1
 Model 4: T-V Picks QC, NMO, Stack

21 tnmo=0.0211082,0.437995,0.844327,0.949868,1.16623,1.98945 \
22 vnmo=1497,1526.7,2551.35,2432.55,2685,2685 \
23 tnmo=0.0105541,0.569921,1.00792,1.30343,1.99472 \
24 vnmo=1497,1541.55,2595.9,2848.35,3011.7 \
25 tnmo=0.0158311,0.686016,1.12401,1.04591,1.98945 \
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

26 vnmo=1482.15,1541.55,2729.55,3160.2,3293.85 \
27 tnmo=0.0158311,0.78628,1.19789,1.5409,1.99472 \
28 vnmo=1482.15,1511.85,3011.7,3694.8,3798.75 \
29 tnmo=0.0105541,0.765172,1.1029,1.45119,1.98945 \
30 vnmo=1497,1511.85,2937.45,3665.1,3858.15 \
31 tnmo=0.0158311,0.627968,0.907652,1.25594,1.99472 \
32 vnmo=1497,1556.4,2595.9,3056.25,3353.25 \
33 tnmo=0.0158311,0.469657,0.686016,1.07124,1.98417 \
34 vnmo=1482.15,1556.4,2521.65,2818.65,3026.55 \
35 tnmo=0.0158311,0.3219,0.501319,0.923483,1.98417 \
36 vnmo=1482.15,1541.55,2462.25,2670.15,2774.1 \
37 tnmo=0.0158311,0.226913,0.395778,0.870712,1.98417 \
38 vnmo=1482.15,1511.85,2521.65,2699.85,2848.35 \
39 tnmo=0.0105541,0.23219,0.390501,0.918206,1.99472 \
40 vnmo=1482.15,1526.7,2566.2,2878.05,3234.45 \
41 tnmo=0.0158311,0.258575,0.46438,1.99472 \
42 vnmo=1497,1526.7,2699.85,3368.1 \
43 tnmo=0.0158311,0.311346,0.538259,0.612137,1.98945 \
44 vnmo=1482.15,1526.7,2759.25,2952.3,3442.35 \
45 tnmo=0.0211082,0.353562,0.622691,0.744063,1.99472 \
46 vnmo=1467.3,1511.85,2774.1,3056.25,3457.2 \
47 tnmo=0.0211082,0.369393,0.691293,0.812665,1.99472 \
48 vnmo=1482.15,1511.85,2878.05,2982,3279 \
49 tnmo=0.0211082,0.348285,0.728232,0.881266,1.20844,1.98945 \
50 vnmo=1482.15,1511.85,2833.5,3026.55,3145.35,3323.55
51
52 # Exit politely from shell
53 exit
54
Before running tvQC.sh, we modified one time value on line 25. Below are the screen
messages from tvQC.sh.
tvnmoqc: This file has 18 CDPs.
tvnmoqc: tnmo values must increase for use in NMO
tvnmoqc: For cdp=160, check times 1.12401 and 1.04591.
tvnmoqc: End of cdp-tnmo-vnmo check.
First, if there had been no time series problems, only the first and last lines would
have printed to the screen. Second, no matter how many time series problems the input
file has, tvnmoqc reads the entire file. It is up to you, the user, to correct time series
problems before you use these lines for NMO (program sunmo). See Section 8.2.3.

8.2.2 Program tvnmoqc, mode = 2

Weeks or months after picking data, we might ask ourselves, “How good (accurate)
were those picks?” The following interactive script, velanQC, answers that question.
Script velanQC makes two images: the seismic CMP and the velan, and it puts the line of
t-v picks on the velan.
Inside script velanQC, we use tvnmoqc with mode=2 and we supply a value for
tvnmoqc parameter prefix. Program tvnmoqc takes each tnmo-vnmo pair, converts the
rows of data to time-velocity (t-v) columns, and writes the columns to a file. The name of
each file is “prefix value” “dot” “cdp value.” For example, the first cdp in file vpick4.txt

8-2
 Model 4: T-V Picks QC, NMO, Stack

is 60. If we set prefix=velqc, the name of the first output file is velqc.60. The
contents of velqc.60 are:
0.0158311 1511.85
0.390501 1511.85
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

0.686016 2640.45
1.06069 2922.6
1.99472 3249.3
To overlay t-v picks on the velan, suximage needs data in columns, not rows.

Figure 8.1: Left: Velan of CMP 60. Right: CMP 60.

Figure 8.2: Left: Velan of CMP 160. Right: CMP 160.

8-3
 Model 4: T-V Picks QC, NMO, Stack

Figure 8.1 shows CMP 60 with the velan and the picks on the velan. Figure 8.2 shows
CMP 160 with the velan and the picks on the velan. The time value that we altered earlier
has not been corrected, so we see the sharp up movement at the altered time. The contents
of file velqc.160 are:
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

0.0158311 1482.15
0.686016 1541.55
1.12401 2729.55
1.04591 3160.2
1.98945 3293.85
As you can see in the file at the end of Chapter 7, the fourth time, 1.04591 should be
1.44591.
Below are the first screen messages from velanQC. Lines numbers are added for
discussion.
1
2 tvnmoqc: This file has 18 CDPs.
3
4 tvnmoqc: tnmo values must increase for use in NMO
5
6 tvnmoqc: For cdp=160, check times 1.12401 and 1.04591.
7
8 tvnmoqc: End of cdp-tnmo-vnmo check.
9
10 *** VELOCITY ANALYSIS QC ***
11
12 Preparing CMP 1 of 18 for Display
13
14 Picks for CMP 60:
15 -------------------------
16 0.0158311 1511.85
17 0.390501 1511.85
18 0.686016 2640.45
19 1.06069 2922.6
20 1.99472 3249.3
21 -------------------------
22
23 Press Return or Enter to see next gather
24 Or enter "x" to exit
25 x
26 Terminated
27 kill: 6548: no such process
28 Terminated
29 kill: 6557: no such process
30
31 ==> Closing velanQC
32
33 press return key to continue
34
Lines 1-8 are the same as are printed when mode=1 because the same checks are
done to the cdp-tnmo-vnmo vales while tvnmoqc reads them. After error-checking and
diagnostic printing, all the files of t-v values are created.
Lines 12-24 are printed as each t-v file is read from disk, and as the CMP and velan
images are made on the screen.
In this example, instead of continuing to the next gather, the user enters “x” to exit the
script: line 25. As a result, messages are printed to the screen as the seismic display

8-4
 Model 4: T-V Picks QC, NMO, Stack

windows are closed (lines 26-29) and the script announces that it is finished processing
(line 31). After the user presses Enter or Return (as instructed on line 33), the script ends.

8.2.2.1 Script velanQC.scr

Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

We usually start interactive scripts by running a .scr script. Here is velanQC.scr:

1 #! /bin/sh
2 # File: velanQC.scr
3 # Run this script to start script velanQC.sh
4
5 xterm -geom 80x20+10+600 -e velanQC.sh
6
Line 1 invokes the shell. Line 5 opens a dialog window 80 characters wide by 20
characters high. Line 5 also places the dialog window 10 pixels from the left side of the
viewing area and 600 pixels from the top of the viewing area. Line 5 also starts
velanQC.sh, the processing script, within that window.

8.2.2.2 Script velanQC.sh

We start interactive scripts by running a .scr script. Here is velanQC.sh.
1 #! /bin/sh
2 # File: velanQC.sh
3 # Run script velanQC.scr to start this script
4
5 # Set messages on
6 ##set -x
7
8 #================================================
9 # USER AREA -- SUPPLY VALUES
10 #------------------------------------------------
11 # CMPs for QC
12
13 cmp1=60 cmp2=80 cmp3=100
14 cmp4=120 cmp5=140 cmp6=160
15 cmp7=180 cmp8=200 cmp9=220
16 cmp10=240 cmp11=260 cmp12=280
17 cmp13=300 cmp14=320 cmp15=340
18 cmp16=360 cmp17=380 cmp18=400
19
20 numCMPs=18
21
22 #------------------------------------------------
23 # Name of files
24 seisdata=../final/cmp4.su # SU seismic
25 prefix=velqc # tvnmoqc cdp-t-v files
26
27 #------------------------------------------------
28 # display choices
29 myperc=98 # perc value for plot
30 plottype=0 # 0 = wiggle plot, 1 = image plot
31
32 #------------------------------------------------
33 # Processing variables
34
35 # Semblance variables
36 nvs=100 # number of velocities
37 dvs=27 # velocity intervals
38 fvs=1200 # first velocity
39
40 #================================================

8-5
 Model 4: T-V Picks QC, NMO, Stack

41 # Insert pick file.

42 # Remove continuation mark (\) from last line.
43
44 tvnmoqc mode=2 prefix=$prefix \
45 cdp=60,80,100,120,140,160,180,200,220,\
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

46 240,260,280,300,320,340,360,380,400 \
47 tnmo=0.0158311,0.390501,0.686016,1.06069,1.99472 \
48 vnmo=1511.85,1511.85,2640.45,2922.6,3249.3 \
49 tnmo=0.0158311,0.353562,0.701847,1.0343,1.99472 \
50 vnmo=1482.15,1511.85,2625.6,2774.1,2982 \
51 tnmo=0.0158311,0.364116,0.765172,0.875989,1.08707,1.98945 \
52 vnmo=1511.85,1526.7,2595.9,2714.7,2625.6,2699.85 \
53 tnmo=0.0211082,0.437995,0.844327,0.949868,1.16623,1.98945 \
54 vnmo=1497,1526.7,2551.35,2432.55,2685,2685 \
55 tnmo=0.0105541,0.569921,1.00792,1.30343,1.99472 \
56 vnmo=1497,1541.55,2595.9,2848.35,3011.7 \
57 tnmo=0.0158311,0.686016,1.12401,1.04591,1.98945 \
58 vnmo=1482.15,1541.55,2729.55,3160.2,3293.85 \
59 tnmo=0.0158311,0.78628,1.19789,1.5409,1.99472 \
60 vnmo=1482.15,1511.85,3011.7,3694.8,3798.75 \
61 tnmo=0.0105541,0.765172,1.1029,1.45119,1.98945 \
62 vnmo=1497,1511.85,2937.45,3665.1,3858.15 \
63 tnmo=0.0158311,0.627968,0.907652,1.25594,1.99472 \
64 vnmo=1497,1556.4,2595.9,3056.25,3353.25 \
65 tnmo=0.0158311,0.469657,0.686016,1.07124,1.98417 \
66 vnmo=1482.15,1556.4,2521.65,2818.65,3026.55 \
67 tnmo=0.0158311,0.3219,0.501319,0.923483,1.98417 \
68 vnmo=1482.15,1541.55,2462.25,2670.15,2774.1 \
69 tnmo=0.0158311,0.226913,0.395778,0.870712,1.98417 \
70 vnmo=1482.15,1511.85,2521.65,2699.85,2848.35 \
71 tnmo=0.0105541,0.23219,0.390501,0.918206,1.99472 \
72 vnmo=1482.15,1526.7,2566.2,2878.05,3234.45 \
73 tnmo=0.0158311,0.258575,0.46438,1.99472 \
74 vnmo=1497,1526.7,2699.85,3368.1 \
75 tnmo=0.0158311,0.311346,0.538259,0.612137,1.98945 \
76 vnmo=1482.15,1526.7,2759.25,2952.3,3442.35 \
77 tnmo=0.0211082,0.353562,0.622691,0.744063,1.99472 \
78 vnmo=1467.3,1511.85,2774.1,3056.25,3457.2 \
79 tnmo=0.0211082,0.369393,0.691293,0.812665,1.99472 \
80 vnmo=1482.15,1511.85,2878.05,2982,3279 \
81 tnmo=0.0211082,0.348285,0.728232,0.881266,1.20844,1.98945 \
82 vnmo=1482.15,1511.85,2833.5,3026.55,3145.35,3323.55
83
84 #================================================
85
86 # HOW SEMBLANCE (VELAN) VELOCITIES ARE COMPUTED
87
88 # Last Vel = fvs + (( nvs-1 ) * dvs ) = lvs
89 # 5000 = 500 + (( 99-1 ) * 45 )
90 # 3900 = 1200 + (( 100-1 ) * 27 )
91
92 # Compute last semblance (velan) velocity
93 lvs=`bc -l << -END
94 $fvs + (( $nvs - 1 ) * $dvs )
95 END`
96
97 #================================================
98
99 # FILE DESCRIPTIONS
100
101 # tmp1 = ASCII temp file to avoid screen display of "zap"
102 # tmp2 = ASCII temp file for "wc" result (velan)
103 # tmp3 = ASCII temp file for stripping file name from tmp2 (velan)

8-6
 Model 4: T-V Picks QC, NMO, Stack

104 # panel.$picknow = current CMP windowed from line of CMPs (seismic)

105 # ${prefix}.${picknow} = current CMP picks arranged as "t1 v1"
106 # "t2 v2"
107 # etc.
108
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

109 #================================================
110
111 echo " "
112 echo " *** VELOCITY ANALYSIS QC ***"
113 echo " "
114
115 #------------------------------------------------
116
117 # Remove old files.
118 rm -f panel.* picks.* tmp*
119
120 #------------------------------------------------
121 # Get ns, dt, first time from seismic file
122 nt=`sugethw ns < $seisdata | sed 1q | sed 's/.*ns=//'`
123 dt=`sugethw dt < $seisdata | sed 1q | sed 's/.*dt=//'`
124 ft=`sugethw delrt < $seisdata | sed 1q | sed 's/.*delrt=//'`
125
126 # Convert dt from header value in microseconds
127 # to seconds for velocity profile plot
128 dt=`bc -l << -END
129 scale=6
130 $dt / 1000000
131 END`
132
133 # If "delrt", use it; else use zero
134 if [ $ft -ne 0 ] ; then
135 tstart=`bc -l << -END
136 scale=6
137 $ft / 1000
138 END`
139 else
140 tstart=0.0
141 fi
142
143 #================================================
144 # BEGIN QC LOOP
145 #------------------------------------------------
146
147 i=1
148 while [ $i -le $numCMPs ]
149 do
150
151 # set variable $picknow to current CMP
152 eval picknow=\$cmp$i
153
154 echo " "
155 echo "Preparing CMP $i of $numCMPs for Display "
156 echo " "
157
158 #------------------------------------------------
159 # Plot CMP (right)
160 #------------------------------------------------
161
162 suwind < $seisdata \
163 key=cdp min=$picknow max=$picknow > panel.$picknow
164
165 if [ $plottype -eq 0 ] ; then
166 suxwigb < panel.$picknow xbox=422 ybox=10 wbox=400 hbox=550 \

8-7
 Model 4: T-V Picks QC, NMO, Stack

167 title="CMP gather $picknow" \

168 label1=" Time (s)" label2="Offset (m)" key=offset \
169 perc=$myperc verbose=0 &
170 else
171 suximage < panel.$picknow xbox=421 ybox=10 wbox=400 hbox=550 \
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

172 title="CMP gather $picknow" \

173 label1=" Time (s)" \
174 perc=$myperc verbose=0 &
175 fi
176
177 #------------------------------------------------
178 # Plot semblance (velan) (left)
179 #------------------------------------------------
180
181 # Get the number of picks (number of lines) in prefix.picknow |
182 # Remove blank spaces preceding the line count.
183 # Remove file name that was returned from "wc".
184 # Store line count in "npair" to guide line on velan.
185
186 echo "Picks for CMP ${picknow}:"
187 echo "-------------------------"
188 cat ${prefix}.${picknow}
189 echo "-------------------------"
190
191 wc -l ${prefix}.${picknow} | sed 's/^ *\(.*\)/\1/' > tmp2
192 sed 's/ ${prefix}.${picknow} //' < tmp2 > tmp3
193 npair=`sort < tmp3`
194
195 suvelan < panel.$picknow nv=$nvs dv=$dvs fv=$fvs |
196 suximage xbox=10 ybox=10 wbox=400 hbox=550 perc=99 \
197 units="semblance" f2=$fvs d2=$dvs n2tic=5 \
198 title="Semblance Plot CMP $picknow" cmap=hsv2 \
199 label1=" Time (s)" label2="Velocity (m/s)" \
200 legend=1 units=Semblance verbose=0 gridcolor=black \
201 grid1=solid grid2=solid \
202 curve=${prefix}.${picknow} npair=$npair curvecolor=white &
203
204 #------------------------------------------------
205 # Question: Next CMP/Velan or Exit?
206 #------------------------------------------------
207
208 echo " "
209 echo "Press Return or Enter to see next gather"
210 echo " Or enter \"x\" to exit"
211 > /dev/tty
212 read response
213
214 case $response in
215 [xX])
216 zap xwigb > tmp1
217 zap ximage > tmp1
218 i=`expr $numCMPs + 1`
219 ;;
220 *)
221 zap xwigb > tmp1
222 zap ximage > tmp1
223 i=`expr $i + 1`
224 ;;
225 esac
226
227 done
228
229 #================================================

8-8
 Model 4: T-V Picks QC, NMO, Stack

230 # Exit
231 #------------------------------------------------
232
233 # Exit
234 echo " "
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

235 echo " ==> Closing velanQC"

236 pause
237
238 rm -f panel.* picks.* tmp*
239 ##rm –f $prefix.*
240 exit
241
Script velanQC.sh is simpler than iva.sh; it only makes two displays and it
continuously moves forward through the gathers. You only choose whether to view the
next gather and velan with its picks or to quit the script.
Program tvnmoqc is called on line 44. The pick file must be placed immediately
after.
Note: Remove the last continuation mark (\) from the pick file once it is inside
velanQC.sh. Set the value for tvnmoqc parameter prefix on line 25.
Unlike iva.sh user parameters, here the user does not supply an outpick file name and
does not supply CVS variables.
The CMPs for QC (lines 13-18) have to be the same CMPs listed in vpick4.txt. You
can omit CMPs, but the ones included must be the same, or the picks will not properly
overlay.
Note: When velanQC.sh finishes, it does not remove the individual t-v files.
However, if you un-comment line 239 (remove “##”), it will remove the files.

8.2.3 Edit Time Picks

What do we do when we find a series with time values that do not increase? Here, we
provide three answers. Before we look at the answers, let’s restate the problem.
Pair 1 shows a tnmo-vnmo series with times and velocities that always increase. We
created this pair in Chapter 7.
Pair 1:
------
25 tnmo=0.0158311,0.686016,1.12401,1.44591,1.98945 \
26 vnmo=1482.15,1541.55,2729.55,3160.2,3293.85 \
Pair 2 shows the tnmo-vnmo series that we used in our examples above. A mistaken
pick like this is possible, but not common. We do not usually move the mouse up as we
make picks down the velan.
Pair 2:
------
25 tnmo=0.0158311,0.686016,1.12401,1.04591,1.98945 \
26 vnmo=1482.15,1541.55,2729.55,3160.2,3293.85 \
Pair 3 shows a more common picking mistake. Here we made two picks on top of
each other. The third and fourth picks are the same – we did not move the mouse between
picks. The result is an extra pick between two intended picks.
Pair 3:
------
25 tnmo=0.0158311,0.686016,1.12401,1.12401,1.44591,1.98945 \
26 vnmo=1482.15,1541.55,2729.55,2729.55,3160.2,3293.85 \

8-9
 Model 4: T-V Picks QC, NMO, Stack

Let’s solve the problem presented by Pair 3.

Our first solution is to remove the extra time-velocity pair. Once we do that, we have
the same tnmo-vnmo series as Pair 2. This is a good solution, but we try not to use it. We
are afraid of making a mistake while typing and backspacing in this crowded file.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Solution 1:
----------
25 tnmo=0.0158311,0.686016,1.12401,1.44591,1.98945 \
26 vnmo=1482.15,1541.55,2729.55,3160.2,3293.85 \
Our second solution is to increase the value of the fourth time by one millisecond (we
do not change the corresponding velocity). We change the fourth time from 1.12401 to
1.12501. This change has just about no effect on NMO, and we spend less time editing
solution 2 than we do solution 1.
Solution 2:
----------
25 tnmo=0.0158311,0.686016,1.12401,1.12501,1.44591,1.98945 \
26 vnmo=1482.15,1541.55,2729.55,2729.55,3160.2,3293.85 \
Our third solution is to re-pick the CMP. Below are lines from iva.sh (Section
7.6.6.3).
10 #------------------------------------------------
11 # CMPs for analysis
12
13 cmp1=60 cmp2=80 cmp3=100
14 cmp4=120 cmp5=140 cmp6=160
15 cmp7=180 cmp8=200 cmp9=220
16 cmp10=240 cmp11=260 cmp12=280
17 cmp13=300 cmp14=320 cmp15=340
18 cmp16=360 cmp17=380 cmp18=400
19
20 numCMPs=18
21
22 #------------------------------------------------
To re-pick CMP 160, we change the lines as shown below. What was “cmp6” is now
“cmp1,” and parameter numCMPs is now “1.” Nothing else in iva.sh needs to change.
10 #------------------------------------------------
11 # CMPs for analysis
12
13
14 cmp1=160
15
16
17
18
19
20 numCMPs=1
21
22 #------------------------------------------------
After we re-pick CMP 160, we replace the old tnmo-vnmo pair with the new tnmo-
vnmo pair.

8.3 Normal Moveout: Script nmo4.sh

The file vpick4.txt that was shown at the end of Chapter 7 is now presented within
script nmo4.sh. As we noted in Section 8.2.1, what was one line in vpick4.txt (line 1) is

8-10
 Model 4: T-V Picks QC, NMO, Stack

now two lines in nmo4.sh (lines 18-19). Also, we removed line 2 from file vpick4.txt
(refer to the end of Chapter 7). That line would be between lines 19 and 20 below.
1 #! /bin/sh
2 # File: nmo4.sh
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

3 # Apply NMO (flatten) 2-D line of CMPs

4 # Input (1): 2-D line of CMPs
5 # Output (1): NMO-corrected 2-D line of CMPs
6 # Use: nmo4.sh
7 #
8 # NMO correction is interpolated between named CMPs.
9
10 # Set debugging on
11 set -x
12
13 # Name data sets
14 indata=cmp4.su
15 outdata=nmo4.su
16
17 sunmo < $indata \
18 cdp=60,80,100,120,140,160,180,200,220,\
19 240,260,280,300,320,340,360,380,400 \
20 tnmo=0.0158311,0.390501,0.686016,1.06069,1.99472 \
21 vnmo=1511.85,1511.85,2640.45,2922.6,3249.3 \
22 tnmo=0.0158311,0.353562,0.701847,1.0343,1.99472 \
23 vnmo=1482.15,1511.85,2625.6,2774.1,2982 \
24 tnmo=0.0158311,0.364116,0.765172,0.875989,1.08707,1.98945 \
25 vnmo=1511.85,1526.7,2595.9,2714.7,2625.6,2699.85 \
26 tnmo=0.0211082,0.437995,0.844327,0.949868,1.16623,1.98945 \
27 vnmo=1497,1526.7,2551.35,2432.55,2685,2685 \
28 tnmo=0.0105541,0.569921,1.00792,1.30343,1.99472 \
29 vnmo=1497,1541.55,2595.9,2848.35,3011.7 \
30 tnmo=0.0158311,0.686016,1.12401,1.44591,1.98945 \
31 vnmo=1482.15,1541.55,2729.55,3160.2,3293.85 \
32 tnmo=0.0158311,0.78628,1.19789,1.5409,1.99472 \
33 vnmo=1482.15,1511.85,3011.7,3694.8,3798.75 \
34 tnmo=0.0105541,0.765172,1.1029,1.45119,1.98945 \
35 vnmo=1497,1511.85,2937.45,3665.1,3858.15 \
36 tnmo=0.0158311,0.627968,0.907652,1.25594,1.99472 \
37 vnmo=1497,1556.4,2595.9,3056.25,3353.25 \
38 tnmo=0.0158311,0.469657,0.686016,1.07124,1.98417 \
39 vnmo=1482.15,1556.4,2521.65,2818.65,3026.55 \
40 tnmo=0.0158311,0.3219,0.501319,0.923483,1.98417 \
41 vnmo=1482.15,1541.55,2462.25,2670.15,2774.1 \
42 tnmo=0.0158311,0.226913,0.395778,0.870712,1.98417 \
43 vnmo=1482.15,1511.85,2521.65,2699.85,2848.35 \
44 tnmo=0.0105541,0.23219,0.390501,0.918206,1.99472 \
45 vnmo=1482.15,1526.7,2566.2,2878.05,3234.45 \
46 tnmo=0.0158311,0.258575,0.46438,1.99472 \
47 vnmo=1497,1526.7,2699.85,3368.1 \
48 tnmo=0.0158311,0.311346,0.538259,0.612137,1.98945 \
49 vnmo=1482.15,1526.7,2759.25,2952.3,3442.35 \
50 tnmo=0.0211082,0.353562,0.622691,0.744063,1.99472 \
51 vnmo=1467.3,1511.85,2774.1,3056.25,3457.2 \
52 tnmo=0.0211082,0.369393,0.691293,0.812665,1.99472 \
53 vnmo=1482.15,1511.85,2878.05,2982,3279 \
54 tnmo=0.0211082,0.348285,0.728232,0.881266,1.20844,1.98945 \
55 vnmo=1482.15,1511.85,2833.5,3026.55,3145.35,3323.55 \
56 > $outdata
57
58 # Exit politely from shell
59 exit
60

8-11
 Model 4: T-V Picks QC, NMO, Stack

This script runs very quickly.

Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 8.3: Left: NMO applied to CMP 120. Center: NMO applied to CMP 130. Right:
NMO applied to CMP 140. For CMP 130, the NMO t-v values were linearly interpolated
between CMP 120 and CMP 140.

Figure 8.4: NMO applied to CMP 130, no interpolation.

8-12
 Model 4: T-V Picks QC, NMO, Stack

Recall that t-v picks were made every 20th CMP starting at the first full-fold CMP,
CMP 60, and ending at the last full-fold CMP, CMP 400 (see Section 7.5). Program
sunmo used the values of script nmo4.sh to interpolate between CMPs.
How accurate is the interpolation? Mathematically, it is quite accurate. How useful is
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

the interpolation? Figure 8.3 shows CMPs 120, 130 and 140 after NMO. We think we
made precise picks to flatten CMPs 120 and 140; their reflectors look reasonably flat. But
the third reflector of CMP 130, with interpolated picks, appears to curve quite a bit.
When we stack CMP 130, we expect that the third reflector will smear; that is, it will not
stack to a zero-offset spike. What is worse, the smear might contribute to smearing the
second reflector. (We are showing the worst CMP of nmo4.su.)
To improve our stack, we modify script iva.sh to pick only CMP 130. After doing so,
we show the flattened gather in Figure 8.4.
This is not perfect flattening, but better than before. Below is the new NMO script,
nmo4a.sh.
1 #! /bin/sh
2 # File: nmo4a.sh
3 # Apply NMO (flatten) 2-D line of CMPs
4 # Input (1): 2-D line of CMPs
5 # Output (1): NMO-corrected 2-D line of CMPs
6 # Use: nmo4a.sh
7 #
8 # NMO correction is interpolated between named CMPs.
9
10 # Set debugging on
11 set -x
12
13 # Name data sets
14 indata=cmp4.su
15 outdata=nmo4a.su
16
17 sunmo < $indata \
18 cdp=60,80,100,120,130,140,160,180,200,220,\
19 240,260,280,300,320,340,360,380,400 \
20 tnmo=0.0158311,0.390501,0.686016,1.06069,1.99472 \
21 vnmo=1511.85,1511.85,2640.45,2922.6,3249.3 \
22 tnmo=0.0158311,0.353562,0.701847,1.0343,1.99472 \
23 vnmo=1482.15,1511.85,2625.6,2774.1,2982 \
24 tnmo=0.0158311,0.364116,0.765172,0.875989,1.08707,1.98945 \
25 vnmo=1511.85,1526.7,2595.9,2714.7,2625.6,2699.85 \
26 tnmo=0.0211082,0.437995,0.844327,0.949868,1.16623,1.98945 \
27 vnmo=1497,1526.7,2551.35,2432.55,2685,2685 \
28 tnmo=0.0211082,0.506596,0.912929,1.05013,1.24011,1.99472 \
29 vnmo=1511.85,1526.7,2595.9,2982,2714.7,3145.35 \
30 tnmo=0.0105541,0.569921,1.00792,1.30343,1.99472 \
31 vnmo=1497,1541.55,2595.9,2848.35,3011.7 \
32 tnmo=0.0158311,0.686016,1.12401,1.44591,1.98945 \
33 vnmo=1482.15,1541.55,2729.55,3160.2,3293.85 \
34 tnmo=0.0158311,0.78628,1.19789,1.5409,1.99472 \
35 vnmo=1482.15,1511.85,3011.7,3694.8,3798.75 \
36 tnmo=0.0105541,0.765172,1.1029,1.45119,1.98945 \
37 vnmo=1497,1511.85,2937.45,3665.1,3858.15 \
38 tnmo=0.0158311,0.627968,0.907652,1.25594,1.99472 \
39 vnmo=1497,1556.4,2595.9,3056.25,3353.25 \
40 tnmo=0.0158311,0.469657,0.686016,1.07124,1.98417 \
41 vnmo=1482.15,1556.4,2521.65,2818.65,3026.55 \
42 tnmo=0.0158311,0.3219,0.501319,0.923483,1.98417 \
43 vnmo=1482.15,1541.55,2462.25,2670.15,2774.1 \

8-13
 Model 4: T-V Picks QC, NMO, Stack

44 tnmo=0.0158311,0.226913,0.395778,0.870712,1.98417 \
45 vnmo=1482.15,1511.85,2521.65,2699.85,2848.35 \
46 tnmo=0.0105541,0.23219,0.390501,0.918206,1.99472 \
47 vnmo=1482.15,1526.7,2566.2,2878.05,3234.45 \
48 tnmo=0.0158311,0.258575,0.46438,1.99472 \
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

49 vnmo=1497,1526.7,2699.85,3368.1 \
50 tnmo=0.0158311,0.311346,0.538259,0.612137,1.98945 \
51 vnmo=1482.15,1526.7,2759.25,2952.3,3442.35 \
52 tnmo=0.0211082,0.353562,0.622691,0.744063,1.99472 \
53 vnmo=1467.3,1511.85,2774.1,3056.25,3457.2 \
54 tnmo=0.0211082,0.369393,0.691293,0.812665,1.99472 \
55 vnmo=1482.15,1511.85,2878.05,2982,3279 \
56 tnmo=0.0211082,0.348285,0.728232,0.881266,1.20844,1.98945 \
57 vnmo=1482.15,1511.85,2833.5,3026.55,3145.35,3323.55 \
58 > $outdata
59
60 # Exit politely from shell
61 exit
62
Script nmo4a.sh is two lines longer than nmo4.sh. Line 18 is modified to include
“130.” The t-v picks for CMP 130 are the fifth tnmo-vnmo pair, lines 28-29.

8.4 Interactive Viewer

We can quickly review the gathers because we have an interactive viewer. Actually,
like velanQC, the only interactivity that scripts iview.scr and iview.sh offer is when the
user chooses to see the next gather or quit the script.

8.4.1 Script iview.scr

We usually start interactive scripts by running a .scr script. Here is iview.scr:
1 #! /bin/sh
2 # File: iview.scr
3 # Run this script to start script iview.sh
4
5 xterm -geom 80x20+10+545 -e iview.sh
6
Line 1 invokes the shell. Line 5 opens a dialog window 80 characters wide by 20
characters high. Line 5 also places the dialog window 10 pixels from the left side of the
viewing area and 545 pixels from the top of the viewing area. Line 5 also starts iview.sh,
the processing script, within that window.

8.4.2 Script iview.sh

We start interactive scripts by running a .scr script. Here is iview.sh.
1 #! /bin/sh
2 # File: iview.sh
3 # View seismic gathers from a 2-D line
4
5 # Set messages on
6 ##set -x
7
8 #================================================
9 # USER AREA -- SUPPLY VALUES
10
11 # input seismic data
12 indata=nmo4a.su # SU format
13
14 # plot choices

8-14
 Model 4: T-V Picks QC, NMO, Stack

15 myperc=98 # perc value for plot

16 plottype=0 # 0 = wiggle plot, 1 = image plot
17 Wplot=400 # Width of plot (pixels)
18 Hplot=500 # Height of plot (pixels)
19
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

20 # processing variables
21 sortkey=cdp # sort key (usually fldr or cdp)
22 firsts=10 # first sort (fldr or cdp) value
23 lasts=458 # last sort (fldr or cdp) value
24 increment=10 # sort key increment
25 tracekey=offset # trace label key
26
27 #================================================
28 # file descriptions
29
30 # tmp1 = binary temp file for input gather
31
32 #------------------------------------------------
33
34 echo " "
35 echo " *** VIEWER ***"
36 echo " "
37 echo " INPUT: $indata"
38
39 #------------------------------------------------
40 # Remove old temporary file
41 rm -f tmp*
42
43 #------------------------------------------------
44 # BEGIN LOOP
45 #------------------------------------------------
46
47 i=$firsts
48 while [ $i -le $lasts ]
49 do
50
51 echo " "
52 echo "Reading gather $i of $indata"
53 echo "First gather = $firsts Last gather = $lasts"
54
55 suwind < $indata key=$sortkey min=$i max=$i > tmp1
56
57 if [ $plottype -eq 0 ] ; then
58 suxwigb < tmp1 xbox=10 ybox=10 wbox=$Wplot hbox=$Hplot \
59 title="$sortkey $i" \
60 label1=" Time (s)" label2="$tracekey" key=$tracekey \
61 perc=$myperc verbose=0 &
62 else
63 suximage < tmp1 xbox=10 ybox=10 wbox=$Wplot hbox=$Hplot \
64 title="$sortkey $i" \
65 label1=" Time (s)" \
66 perc=$myperc verbose=0 &
67 fi
68
69 echo " "
70 echo "Press Return or Enter to see next gather"
71 echo " Or enter \"x\" to exit"
72 > /dev/tty
73 read response
74
75 case $response in
76 [xX])
77 zap xwigb > tmp1

8-15
 Model 4: T-V Picks QC, NMO, Stack

78 zap ximage > tmp1

79 i=`expr $lasts + 1`
80 ;;
81 *)
82 zap xwigb > tmp1
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

83 zap ximage > tmp1

84 i=`expr $i + $increment`
85 ;;
86 esac
87
88 done
89
90 #------------------------------------------------
91 # END LOOP
92 #------------------------------------------------
93
94 # Exit
95 echo " "
96 echo " ==> Closing iview"
97 pause
98 rm -f tmp*
99 exit
100
The user-supplied values are on lines 8-26.
The gather loop is lines 43-92. Line 48 tests whether the value of “i” is within the
processing range. If we decide to exit the script before viewing all gathers (by entering
“x”), line 79 sets the value of “i” to one larger than variable lasts to make the script jump
out of the view loop.
There are only two steps to using the script:
1. Click in the dialog window to make it active.
2. Press Return or Enter to see the next gather or enter “X” to exit the viewer.
Following is the screen dialogue of the first two gathers and the last two gathers
(excluding system messages):
*** VIEWER ***

INPUT: nmo4a.su

Reading gather 10 of nmo4a.su

First gather = 10 Last gather = 458

Press Return or Enter to see next gather

Or enter "x" to exit

Reading gather 20 of nmo4a.su

First gather = 10 Last gather = 458

Press Return or Enter to see next gather

Or enter "x" to exit

...

Reading gather 440 of nmo4a.su

First gather = 10 Last gather = 458

Press Return or Enter to see next gather

Or enter "x" to exit

8-16
 Model 4: T-V Picks QC, NMO, Stack

Reading gather 450 of nmo4a.su

First gather = 10 Last gather = 458

Press Return or Enter to see next gather

Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Or enter "x" to exit

==> Closing iview

press return key to continue

8.5 Stack
A one-line command is sufficient to stack the data. The default stack key is cdp.
sustack < nmo4a.su > stack4.su
Below is the surange output of file stack4.su.
surange < stack4.su
458 traces:
tracl=(1,458) tracr=(1,12000) fldr=(1,200) tracf=(1,60) cdp=(1,458)
trid=1 nhs=(1,30) sx=(0,9950) gx=(-1475,11425) ns=501
dt=4000

8-17
 Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

8-18
Model 4: T-V Picks QC, NMO, Stack

Figure 8.5: An ximage plot of stack4.su. The horizontal axis is CMP number. The sand channel diffraction is near CMP 110, at
approximately 0.8 seconds.
 Model 4: T-V Picks QC, NMO, Stack
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 8.6: Arrows point to velocity analysis CMP locations. Between those CMPs, the
top reflector thickens.
When we look at steep portions of the top reflector (Figure 8.6), we see that at CMP
locations where we did velocity analysis, the top reflector is thin. We might say it is well
focused. Between velocity analysis CMPs where sunmo interpolated t-v values, the top
reflector is broader (out of focus, diffuse). The lesson: The steeper the reflector, the more
often velocity analysis should be done.
Another lesson (Figure 8.7): Velocity analysis should be done more frequently where
velocity changes rapidly (vertically or laterally). The channel on the left side of the stack,
the diffraction with an apex near CMP 110 at approximately 0.8 seconds, is such a
velocity discontinuity.

Figure 8.7: The arrow points to a diffraction hyperbola, the time section expression of
the sand channel.

8-19
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

This page has been intentionally left blank

 Model 4: Migration

9. Model 4: Migration
9.1 Introduction
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

In this chapter, we develop tools that migrate data and that display migrated data.

9.2 Migration Constant Velocity Panels (CVP) with Stolt Migration

9.2.1 sustolt Parameters

In Section 9.2.2, we will use migration program sustolt. Here, we examine its
parameters (except verbose and tmpdir).
Table 9.1: Processing parameters of sustolt (R=required, O=optional)
Name Default Descrption
cdpmin R minimum CDP number for which to apply DMO
cdpmax R maximum CDP number for which to apply DMO
dxcdp R distance between adjacent CDP bins (m)
noffmix O 1 number of offsets to mix (for unstacked data only)
tmig O 0.0 times corresponding to RMS velocities in vmig (s)
vmig O 1500. RMS velocities corresponding to times in tmig (m/s)
smig O 1.0 stretch factor (0.6 typical if Vrms increasing)
vscale O 1.0 scale factor to apply to all velocities
fmax O Nyquist maximum frequency in input data (Hz)
lstaper O 0 length of side tapers (traces)
lbtaper O 0 length of bottom taper (samples)
The first two parameters, cdpmin and cdpmax, are required and have no default
values. They allow us to apply DMO to a specified range of input CDPs. We will always
set these minimum and maximum values to include all our input CDPs. In other words,
we will always let sustolt apply DMO to all input CDPs. DMO, dip moveout correction,
is NMO corrected for dipping beds in prestack data. Because we are migrating stacked
traces, range refers to the CDPs that will be migrated.
The third parameter, dxcdp, is required and has no default value. To determine the
CDP bin distance of Model 4, we created a stacking chart similar to the one shown in
Sheriff (2002). Figure 9.1 is a subset of the acquisition geometry described in Section 6.3,
but it includes enough geometry to let us calculate the CDP bin distance. Model 4 was
acquired with a split-spread line of 60 receivers, receivers spaced 50 meters apart, with
line moves 50 meters between shots. The split-spread line of Figure 9.1 has only 10
receivers, but receiver spacing and line movement are the same as for Model 4. The
diagonal line of Common Midpoint Direction cuts the diagonal of 50 meter x 50 meter
boxes. Therefore,
Common Midpoint Distance = ¥( 50x50 + 50x50 ) = 70.71 m

9-1
 Model 4: Migration

Since we will migrate stacked data, we will ignore parameter noffmix.

We will discuss parameters tmig and vmig, migration time and velocity pairs, later.
We will set the stretch factor, parameter smig
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

• to 1.0 when we do constant velocity migration or

• to 0.6 when we do time-varying migration.
Sometimes we find an SU parameter that defies our best efforts to understand it.
Parameter vscale is such a parameter. When we can calibrate the migration against a
model, as we can for Model 4, we determine the velocity scale factor, vscale,
experimentally (see the next section). When we do not have a model, we use vscale=1.0.

Figure 9.1: A reduced stacking chart of Model 4. Circles are receivers, squares are
sources.
If we supply a frequency value less than the Nyquist for fmax, we can reduce the
number of frequencies that sustolt migrates. For example, the Nyquist frequency of
Model 4 (4 ms sample interval) is 125 Hz. If we decide we do not have usable signal
above 75 Hz, we could use fmax=75 and sustolt will migrate 75 frequencies instead of
125 frequencies; a saving of almost 50%. However, we are pleased with the speed of
sustolt, so we will not use parameter fmax.
We experimented with the length of the taper on the sides, lstaper, and the length of
the taper on the bottom, lbtaper. We decided to use lstaper=20 (traces) and lbtaper=100
(samples). We encourage you to make your own tests.

9.2.2 Seismic Migration with Model 4 and sustolt

Migration is a process by which reflectors are re-positioned to their true subsurface
locations. Frequently, migration is also used to obtain subsurface velocities. That is,
velocities used in migration that collapse diffractions and otherwise properly re-position
reflectors are considered the rms (root-mean-square) velocities of the subsurface layers.

9-2
 Model 4: Migration

The sfill values we used to build Model 4 (Section 6.2) are the rms velocities. Since
we know the rms velocities, we will migrate a portion of Model 4, the sand channel
diffractor, to determine the correct value of sustolt parameter vscale. Once we know the
value of vscale that collapses the sand channel diffraction using a velocity of
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

approximately 3000 m/s, we will migrate the entire model.

9.2.3 Migration Constant Velocity Panels (CVPs)

We use migration program sustolt because it is fast. Here, we use it to create a series
of constant velocity migrated sections. Script migcvp.sh takes in one stack section,
migrates it several times, each time with a constant velocity. We call each migration
product a “panel.” All panels are together in the same output file (Figure 9.2). After
migration, we use iviewcvp.scr and iviewcvp.sh (Section 9.3), variations of iview.scr and
iview.sh (Section 8.4) to review the migration panels.
When we used surange on stack4.su (Section 8.5), we learned that the range of cdp
values is 1-458. We can supply these values in migcvp.sh for variables cdpmin and
cdpmax (lines 20-21). However, let us first examine the sand channel diffraction near
CMP 110. The sand channel is a small event for which we know the model velocity:
3015 m/s (Table 6.2, Section 6.2). We can use the model velocity to determine sustolt
parameter vscale.

Figure 9.2: When the input to script migcvp.sh contains 50 traces and migcvp.sh uses six
different velocities, the output file contains 300 traces.
We use lines 20 and 21 to select a window of CMPs around CMP 110; we migrate
CMPs 80-150. Line 22 uses our calculated value of CDP bin spacing.
Since migcvp.sh does constant velocity migration
• we supplied 1.0 for smig, the stretch factor (line 23).
• we supplied “0” for tmig (line 79).
What should we use for the value of vscale? After several tests, we found that a
velocity scale factor of 1.9 (line 25) yielded the best migration of the sand channel. We
determined vscale by repeatedly migrating the sand channel diffraction until the best
migration image was obtained for a migration velocity of 3000 m/s, the value we know to
be the true interval velocity. Figure 9.3 was obtained using vscale=1.5 and Figure 9.4
was obtained using vscale=1.9.
(In seismic data processing, we sometimes have to know the answer to get the right
answer!)
As discussed in the previous section, the length of the taper on each side (lstaper) is
set to 20 traces (line 26) and the length of the bottom taper (lbtaper) is set to 100 samples
(line 27).

9-3
 Model 4: Migration

Lines 30-32 specify that the first panel is migrated at 1200 m/s, the last panel is
migrated at 5000 m/s, and the panel interval is 200 m/s. These numbers mean migcvp.sh
migrates the input data 20 times (see line 41):
( (5000-1200) / 200 ) + 1 = 20
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Line 34 is for testing. Set variable numVtest to “1” or “2” or another small value to
limit the number of panels processed (lines 45-47).
1 #! /bin/sh
2 # File: migcvp.sh
3 # Create one panel for each migration velocity
4 # Each panel has the same "fldr" value
5 # The migration velocity is in key "offset"
6 # Total number of panels is in key "nvs"
7
8 # Set messages on
9 ##set -x
10
11 #================================================
12 # USER AREA -- SUPPLY VALUES
13 #------------------------------------------------
14
15 # Seismic files
16 indata=stack4.su # SU format
17 outdata=migcvp.su # migration Constant Velocity Panels
18
19 # Migration variables
20 cdpmin=80 # Start CDP value
21 cdpmax=150 # End CDP value
22 dxcdp=70.71 # distance between adjacent CDP bins (m)
23 smig=1.0 # stretch factor (0.6 typical if vrms increasing)
24 # [the "W" factor] (Default=1.0)
25 vscale=1.9 # scale factor to apply to velocities (Default=1.0)
26 lstaper=20 # length of side tapers (traces) (Default=0)
27 lbtaper=100 # length of bottom taper (samples) (Default=0)
28
29 # Velocity panel variables
30 firstv=1200 # first velocity value
31 lastv=5000 # last velocity value
32 increment=200 # velocity increment
33
34 numVtest=100 # use to limit number of velocity panels
35 # otherwise, use very large value (100)
36
37 #================================================
38
39 # Compute number of velocity panels
40
41 numV=`bc -l << -END
42 ( ( $lastv - $firstv ) / $increment ) + 1
43 END`
44
45 if [ $numVtest -lt $numV ] ; then
46 numV=$numVtest
47 fi
48
49 #------------------------------------------------
50
51 # FILE DESCRIPTIONS
52 # tmp1 = binary temp file of input data
53
54 #------------------------------------------------

9-4
 Model 4: Migration

55
56 cp $indata tmp1
57 migV=$firstv
58 echo " "
59
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

60 #------------------------------------------------
61 # Loop through Migration Constant Velocity Panels
62 # Each panel has the same "fldr" value
63 # Panel migration velocity is in key "offset"
64 # Total number of panels (numV) is in key "nvs"
65 #------------------------------------------------
66
67 i=1
68 while [ $i -le $numV ]
69 do
70
71 echo " iteration number = $i Velocity = $migV"
72
73 suwind < tmp1 key=cdp min=$cdpmin max=$cdpmax |
74 sushw key=fldr a=$i |
75 sushw key=offset a=$migV |
76 sushw key=nvs a=$numV |
77
78 sustolt cdpmin=$cdpmin cdpmax=$cdpmax dxcdp=$dxcdp \
79 tmig=0 vmig=$migV smig=$smig vscale=$vscale \
80 lstaper=$lstaper lbtaper=$lbtaper \
81 >> $outdata
82
83 i=`expr $i + 1`
84 migV=`expr $migV + $increment`
85
86 done
87
88 #------------------------------------------------
89 # Remove files and exit
90 #------------------------------------------------
91
92 echo " "
93 echo " Output file = $outdata"
94 echo " "
95
96 rm -f tmp*
97 exit
98
Line 56 copies the input data to a temporary file (tmp1) to avoid changing the original
data. Line 57 sets the first migration velocity.
The loop is lines 60-87. Line 71 writes a message to the screen: the current migration
panel number and its migration velocity. Line 73 windows the appropriate CMPs from
the temporary file. Line 74 sets key fldr to the iteration counter i for all traces in a panel.
Line 75 puts the migration velocity in key offset for all traces in a panel. Line 76 puts the
number of panels in key nvs of every trace. Stolt migration is done in lines 78-80. Line 81
concatenates the output panels. Line 83 increments the loop (and panel) counter. Line 84
increments the velocity for the next panel.
Below are the screen messages for this run:
iteration number = 1 Velocity = 1200
iteration number = 2 Velocity = 1400
iteration number = 3 Velocity = 1600
iteration number = 4 Velocity = 1800

9-5
 Model 4: Migration

iteration number = 5 Velocity = 2000

iteration number = 6 Velocity = 2200
iteration number = 7 Velocity = 2400
iteration number = 8 Velocity = 2600
iteration number = 9 Velocity = 2800
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

iteration number = 10 Velocity = 3000

iteration number = 11 Velocity = 3200
iteration number = 12 Velocity = 3400
iteration number = 13 Velocity = 3600
iteration number = 14 Velocity = 3800
iteration number = 15 Velocity = 4000
iteration number = 16 Velocity = 4200
iteration number = 17 Velocity = 4400
iteration number = 18 Velocity = 4600
iteration number = 19 Velocity = 4800
iteration number = 20 Velocity = 5000

Output file = migcvp.su

9.3 Interactive Constant Velocity Panel Viewer

To view the migration CVPs, we modified the interactive viewer that we used in
Section 8.4. We now have scripts iviewcvp.scr and iviewcvp.sh. Before we look at the
scripts, let’s look at the keys (headers) of the output of migcvp.sh.
surange < migcvp.su
1420 traces:
tracl=(80,150) tracr=(1501,3601) fldr=(1,20) tracf=(1,2) cdp=(80,150)
trid=1 nvs=20 nhs=30 offset=(1200,5000) sx=(1950,3700)
gx=(525,2275) ns=501 dt=4000
As expected, the range of fldr values is 1-20, there are 1420 traces (20 panels x 71
stack traces (cdp=80-150)), and the range of offset values (migration velocities) is 1200-
5000. Last, key nvs has a single value for all traces (20), the number of panels.

9.3.1 Script iviewcvp.scr

We usually start interactive scripts by running a .scr script. Here is iviewcvp.scr:
1 #! /bin/sh
2 # File: iviewcvp.scr
3 # Run this script to start script iviewcvp.sh
4
5 xterm -geom 80x20+10+545 -e iviewcvp.sh
6
Line 1 invokes the shell. Line 5 opens a dialog window 80 characters wide by 20
characters high. Line 5 also places the dialog window 10 pixels from the left side of the
viewing area and 545 pixels from the top of the viewing area. Line 5 also starts
iviewcvp.sh, the processing script, within that window.

9.3.2 Script iviewcvp.sh

We start interactive scripts by running a .scr script. Here is iviewcvp.sh.
The documentation at the top of the script lists important key information:
• key fldr numbers the panels
• key offset contains each panel’s migration velocity
• key nvs contains the total number of panels

9-6
 Model 4: Migration

Because these keys contain these values, and because the panel number and migration
velocity is displayed in each plot, we don’t have to remember migcvp.sh processing
parameters.
Use line 16 to supply the input file name and lines 19-22 to set your plot preferences.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Default settings: The panels are displayed from the first (line 27) to the last. Line 28
is a comment line – the actual last panel value comes from key nvs, retrieved by line 52.
Panel display increment is “1” (line 29), and the x-axis label is key cdp (line 30).
1 #! /bin/sh
2 # File: iviewcvp.sh
3 # View seismic panels from a migration line of
4 # Constant Velocity Panels -- key fldr
5 # Migration velocity of each panel is in key offset
6 # Total number of panels is in key nvs
7
8 # Set messages on
9 ##set -x
10
11 #================================================
12 # USER AREA -- SUPPLY VALUES
13 #------------------------------------------------
14
15 # Input seismic data
16 indata=migcvp.su # SU format
17
18 # Plot choices
19 myperc=98 # perc value for plot
20 plottype=1 # 0 = wiggle plot, 1 = image plot
21 Wplot=300 # Width of plot (pixels)
22 Hplot=500 # Height of plot (pixels)
23
24 #================================================
25 # Processing variables
26 sortkey=fldr # sort key [Do Not Change]
27 firsts=1 # first sort (fldr) value:
28 # lasts # last sort (fldr) value: key nvs
29 increment=1 # sort key (fldr) increment
30 tracekey=cdp # trace label key
31
32 #------------------------------------------------
33 # file descriptions
34
35 # tmp1 = binary file of input panel
36 # tmp2 = ASCII file to reduce "zap" screen messages
37
38 #------------------------------------------------
39
40 echo " "
41 echo " *** MIGRATION CONSTANT VELOCITY PANEL VIEWER ***"
42 echo " "
43 echo " INPUT: $indata"
44
45 #------------------------------------------------
46 # Remove old temporary files
47 rm -f tmp*
48
49 #------------------------------------------------
50 # Get first trace - total number of panels in key nvs
51
52 lasts=`sugethw nvs < $indata | sed 1q | sed 's/.*nvs=//'`
53

9-7
 Model 4: Migration

54 #------------------------------------------------
55 # BEGIN PANEL LOOP
56 #------------------------------------------------
57
58 i=$firsts
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

59 while [ $i -le $lasts ]

60 do
61
62 suwind < $indata key=$sortkey min=$i max=$i > tmp1
63
64 # Retrieve migration velocity from key offset of first trace
65 migV=`sugethw offset < tmp1 | sed 1q | sed 's/.*offset=//'`
66 migVel=`expr $migV` # Ensure migration velocity is numeric
67
68 echo " "
69 echo "Reading panel $i of $indata, Velocity $migVel"
70 echo "First panel = $firsts Last panel = $lasts"
71
72 if [ $plottype -eq 0 ] ; then
73 suxwigb < tmp1 xbox=10 ybox=10 wbox=$Wplot hbox=$Hplot \
74 title="$sortkey $i Velocity $migVel" \
75 label1=" Time (s)" label2="$tracekey" key=$tracekey \
76 perc=$myperc verbose=0 &
77 else
78 suximage < tmp1 xbox=10 ybox=10 wbox=$Wplot hbox=$Hplot \
79 title="$sortkey $i Velocity $migVel" \
80 label1=" Time (s)" label2="$tracekey" key=$tracekey \
81 perc=$myperc verbose=0 &
82 fi
83
84 echo " "
85 echo "Press Return or Enter to see next panel"
86 echo " Or enter \"x\" to exit"
87 > /dev/tty
88 read response
89
90 case $response in
91 [xX])
92 zap xwigb > tmp2
93 zap ximage > tmp2
94 i=`expr $lasts + 1`
95 ;;
96 *)
97 zap xwigb > tmp2
98 zap ximage > tmp2
99 i=`expr $i + $increment`
100 ;;
101 esac
102
103 done
104
105 #------------------------------------------------
106 # END PANEL LOOP
107 #------------------------------------------------
108
109 # Exit
110 echo " "
111 echo " ==> Closing iview"
112 pause
113 rm -f tmp*
114 exit
115

9-8
 Model 4: Migration

Let’s look at panel loop lines 54-104.

We want to get the offset value, the panel’s migration velocity, so we can print it to
the screen and put it on the plot. Line 65 is complex; it gets the offset value from the first
trace of file tmp1. The syntax of this line can be used to extract the value of any key from
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

the first trace of tmp1. (We saw this syntax in Section 7.6.6 in script iva.sh.) Line 62
windows the “current” panel to ensure that line 65 extracts the correct velocity from key
offset.
Line 66 uses “expr” to ensure that what is extracted from the trace header is purely
numeric. In Figure 9.3, the migration velocity printed on the plot includes a strange
character. When we include line 66, we do not get the strange character (Figure 9.4).

Figure 9.3: Images from iviewcvp.sh using vscale=1.5. Left: Migration velocity 3400
m/s is too small to collapse the sand channel diffraction. The diffraction is under-
migrated. Right: Migration velocity 4200 m/s is too large. Due to over-migration, the

9-9
 Model 4: Migration

diffraction is now a “smile.” The strange character after “3400” and “4200” is
discussed in the text.
Remember, suximage reads the first tracekey value (here it is cdp) and increments by
“1” (see Section 7.7). If tracekey values do not increment by “1” (for example, due to
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

missing CMPs), the x-axis annotation will be wrong.

In Figure 9.3, notice the many migration “smiles” of the water bottom (1500 m/s rms
velocity). These are the result of overmigration (the migration velocity is too high). These
figures are designed to find the best value for vscale (velocity scale factor) of the deeper
sand channel diffraction. (3015 m/s). After using migcvp.sh and the viewer, we are
satisfied with vscale=1.9 (Figure 9.4).

Figure 9.4: vscale=1.9. Left: Migration velocity for this panel, 2600 m/s, under-
migrates the sand channel. Right: Migration velocity for this panel, 3400 m/s, over-
migrates the sand channel.

9-10
 Model 4: Migration

9.4 Migration Movie

The following script runs a movie of the migration panels. Program suxmovie (x-
movie plot of a segy data set) and parameter loop are the core of the script. We wrote the
parameter explanations, lines 12-25, in terms of the migration constant velocity panels.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

The definitions of these parameters are slightly different in the suxmovie and xmovie
documentation. The following are lines from the xmovie selfdoc:
Button 1 Zoom with rubberband box
Button 2 reverse the direction of the movie.
Button 3 stop and start the movie.
q or Q key Quit
s or S key stop display and switch to Step mode
b or B key set frame direction to Backward
f or F key set frame direction to Forward
n or N key same as 'f'
c or C key set display mode to Continuous mode
If you switch from Movie mode to Step mode by pressing the S key, you can step
forward with the F key or backward with the B key. To restart the movie, press the C key.
1 #! /bin/sh
2 # File: migmovie.sh
3 # Run a "movie" of the migration panels
4 # Plot "title" shows panel velocity
5 # Enter "xmovie" for mouse and keyboard options
6
7 set -x
8
9 indata=migcvp.su
10 perc=98
11
12 loop=1 # run panels forward continuously
13 # 2 = run panels back and forth continuously
14 # 0 = load all panels then stop
15
16 n1=501 # number of time samples
17 d1=0.004 # time sample interval
18 n2=71 # number of traces per panel
19 d2=1 # trace spacing
20
21 width=300 # width of window
22 height=500 # height of window
23
24 fframe=1200 # velocity of first panel for title annotation
25 dframe=200 # panel velocity increment for title annotation
26
27 suxmovie < $indata perc=$perc loop=$loop \
28 n1=$n1 d1=$d1 n2=$n2 d2=$d2 \
29 width=$width height=$height \
30 fframe=$fframe dframe=$dframe \
31 title="Velocity %g" &
32
33 exit
34
Parameters fframe and dframe (lines 24-25) are used to calculate a value for each
panel. As the panel (frame) increments, the value of dframe is incrementally added to
fframe. On line 31, “%g” prints the calculated value in the title. If you do not use
parameters fframe and dframe, you can use “%g” to print the frame number in the title.
Observe that the sand channel moves updip as the migration velocity increases.

9-11
 Model 4: Migration

9.5 Migrate Model 4

The important migration parameters of sustolt are dxcdp, tmig, vmig, smig, and
vscale. We calculated dxcdp from the stacking chart. Previously, we used sustolt as a
constant velocity migration program. For that, tmig was unimportant, vmig had a single
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

value, and smig was 1.0 (unused). We used migcvp.sh to determine a value for vscale
based on the known modeled velocity of the sand channel.
Here, we use sustolt for 1-D (time-varying) migration in script migStolt.sh. Our script
has a single time-varying velocity function that is applied across the section, regardless of
bed dip. We designed our t-v function to fit the data best near CMP 110, the location of
the sand channel diffraction. We selected velocities at different times based on the panels
of constant-velocity migrations. We found that a migration velocity at 1500 m/s was
appropriate at 0.70 seconds; greater migration velocities provided improved images at
later times.
Our migration script, migStolt.sh, does not output the migrated data, just a plot
(Figure 9.5). By now, you know how to modify the script to produce a data file and make
a plot from that data (for example, lines 13-19 of showshot.sh, Section 5.6.)
1 #! /bin/sh
2 # File: migStolt.sh
3 # Stolt migration stacked data
4 # Input: stack data
5 # Output: plot of migrated data
6 # Use: migStolt.sh
7 # Example: migStolt.sh
8
9 # Velocities are stacking (Vrms)
10 # Here, we use the false assumption that stacking
11 # and migrating velocities do not change laterally.
12
13 # smig = stretch factor (0.6 typical if vrms increasing)
14 # vscale = scale factor to apply to velocities
15
16 # Set messages on
17 set -x
18
19 time=0.00,0.70,0.75,0.80,0.85,0.90,2.00
20 vels=1500,1500,2500,3000,2800,3100,3500
21
22 sustolt < stack4.su \
23 cdpmin=1 cdpmax=458 dxcdp=70.71 \
24 tmig=$time vmig=$vels \
25 smig=0.6 vscale=1.9 lstaper=20 lbtaper=100 |
26
27 suximage xbox=10 ybox=10 wbox=800 hbox=400 \
28 title="Migration: Stolt T = $time V = $vels" \
29 label1=" Time (s)" label2="CMP" key=cdp \
30 perc=99 verbose=0 &
31
32 # Exit politely from shell
33 exit
34
The difference between our two uses of sustolt (aside from our use of t-v pairs here)
is that we set smig=0.6 because we are using increasing RMS velocities.
The kink in the second interface at approximately CMP 240 is from the velocity
change from 1500 m/s to 2500 m/s at 0.7 seconds. This velocity change works well for

9-12
 Model 4: Migration

the sand channel, but not well at CMP 240. A time-varying and spatially-varying
migration would improve the image.
In the pre-migration stack section (Figure 8.3), the sand channel is a strong diffraction
and the lower synclines near CMP 180 are unrealistically narrow. After migration (Figure
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

9.5), the sand channel is greatly collapsed and the lower synclines are realistically gentle.
Careful velocity analysis around CMP 110 (not shown) at the diffractor time of 0.84
seconds produced a stacking velocity of 2670 m/s. It is generally accepted that migration
velocities are approximately 10% greater than stacking velocities. 3015 m/s is 13%
greater than 2670 m/s.

9-13
 Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

9-14
Model 4: Migration

Figure 9.5: An ximage plot of stack4.su after Stolt migration. Parameter vscale=1.9. The horizontal axis is CMP number. The
sand channel diffraction near CMP 110, at approximately 0.8 seconds is greatly reduced (collapsed). Compare to Figure 8.3.
 Nankai Data: Examine, Resample, Sort, Gain

10. Nankai Data: Examine, Resample, Sort, Gain

10.1 Introduction
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

We are going to process a real 2-D line of seismic data provided by Prof. Greg Moore
of the University of Hawaii. The data were collected near the coast of Japan, over the
Nankai trough, where the Philippines plate is subducting beneath Eurasia. The “Nankai”
data were collected by the University of Texas, the University of Tulsa, and the
University of Tokyo. Based on this data set, a paper was published (Moore, et al., 1990)
in which this line is called NT62-8.
The following Nankai seismic files are available:
Table 10.1: Nankai seismic data
Bytes Name Description Traces Gathers
423827680 Nshots.su shot-ordered gathers 19057 326
26509560 Nstack.su stacked data, as published 2869 1

10.1.1 Nankai Shot Gathers

Below is the surange output of Nshots.su.
surange < Nshots.su
19057 traces:
tracl=(39069,58125) tracr=(1,19057) fldr=(1687,2012) tracf=(28,96)
cdp=(900,1300)
cdpt=(1,69) trid=(1,2) offset=(-2435,-170) scalel=-10000 scalco=-10000
counit=1 muts=(0,11000) ns=5500 dt=2000 day=206
hour=(21,22) minute=(0,59) sec=(1,59)
Table 10.2: Some Nshots.su key values
Key Range
fldr 1687 to 2012
cdp 900 to 1300
offset -2435 to -170
ns 5500
dt 2000
Keys tracl and tracr number the 19057 traces (although they use different starting
numbers).
Key fldr tells us there are 326 shot gathers (2012-1687+1).
Since the number of samples per trace (ns) is 5500 and the trace sample interval (dt)
is 2000 microseconds, the trace length is 11 seconds (5500 samples/trace x 0.002
seconds/sample)
Notice that the Nankai shot gather trace headers already contain cdp values.
Below is a chart of the number of traces per shot gather (Figure 10.1). (We describe
the program, sukeycount, which generated the data for the figure in Appendix C.) The
first shot gather, 1687, has one trace, then the number of traces per gather increases to a
maximum of 69 at gather 1735. This is constant through gather 1965. After gather 1965,
the number of traces per gather steadily decreases to 1 at the last gather, 2012.

10-1
 Nankai Data: Examine, Resample, Sort, Gain
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 10.1: Traces per shot gather in data set Nshots.su.

10.1.2 Nankai Stack File

Below is the surange output of Nstack.su. This stack file (Figure 10.2) is the product
of processing leading to publication by the University of Hawaii.
surange < Nstack.su
2869 traces:
tracl=(1,2869) fldr=(0,3383) tracf=(0,96) ep=(69,2899) cdp=(1,2869)
cdpt=1 trid=1 nhs=(0,72) offset=(-870,0) ns=2250
dt=4000 nofils=(281,3149) lcf=(69,2899) hcf=1 lcs=(0,3383)
hcs=(0,96) year=(0,207) hour=(0,23) minute=(0,59) sec=(0,59)
grnors=(0,72)
The file of shot gathers, Nshots.su contains 401 CMPs (the range of cdp values is
900-1300); however, the stack file, Nstack.su, contains 2869 traces; in other words,
Nstack.su is the stack output of 2869 CMPs. This tells us that the shot gather file is a
subset of the file that was used by the original processors to make the published stack
section.
Table 10.3: Some Nstack.su key values
Key Range
cdp 1 to 2869
ns 2250
dt 4000
File Nshots.su has 5500 samples with 2 ms sample interval: 11 seconds of data. In
contrast, Nstack.su has 2250 traces with 4 ms sample interval: 9 seconds of data. Last,
notice that these data were collected in very deep water, approximately 6 seconds two-
way time ( (6/2) seconds x 1500 meters/second = 4.5 km ).

10-2
 Nankai Data: Examine, Resample, Sort, Gain
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 10.2: Stack section Nstack.su as used for publication by Moore et al., 1990.

10.1.3 Our Task

After we examine the shot gathers, we will follow the example of the published data
set – we will resample the file from 2 ms to 4 ms sample interval to reduce our prestack
processing load by half. Although there does not appear to be any reflection character

10-3
 Nankai Data: Examine, Resample, Sort, Gain

below eight seconds (see Figure 10.3), we will use all 11 seconds to ensure that we
include diffraction limbs for migration.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 10.3: Two views of Nstack.su. Left: CMPs 900-1300. Right: CMPs 900-1300,
time 5.5-8 seconds.
Remember that on CMP gathers, shallow reflections generally exhibit large moveout
(large curvature) and deep reflections exhibit little moveout (little curvature). The Nankai
data were collected in deep water, so we do not expect to see much moveout on the CMP
gathers. Lack of moveout will make velocity analysis difficult, but the quality of the stack
is not strongly dependent on the exact velocities used beneath the water bottom.

10.2 Computer Note 12: Symbolic Link

We think these files are too large for everyone to have his or her own copy.
Fortunately, there is a simple way for everyone in a group to refer to one copy of the
Nankai data: create a symbolic link to the data directory. Suppose the Nankai files are in
directory
/home/uofhawaii/nankai/Ndata
and I (Forel) am in directory
/home/forel/suscripts
Here, I enter command
ln –s /home/uofhawaii/nankai/Ndata .
Notice the period at the end. The period means “here” (see Section 6.6). You can read
this command as, “Make a symbolic link from … to …
ln –s from to

10-4
 Nankai Data: Examine, Resample, Sort, Gain

This command has the effect of placing a virtual subdirectory (Ndata) in my directory
without actually placing the large data files there. From my suscripts directory, I can
view a listing of the contents of directory Ndata by typing
/home/forel/suscripts> ls -lF Ndata
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

total 160
-r--r--r-- 1 pennin geofac 4238276805 May 22 16:46 Nshots.su
-r--r--r-- 1 pennin geofac 26509560 May 22 16:44 Nstack.su
Note: To prevent accidentally deleting the data, we recommend you set permissions
so the Nankai files are only readable. A brief discussion of permissions is in Section
2.3.5. We also recommend that you store copies of these files in another directory or on a
DVD so you can quickly replace deleted files.

10.3 View the Line of Shot Gathers

Before we begin processing the Nankai data, we should examine some of the shot
gathers. This is a simple but important quality control (QC) step. Nothing is more
important than actually looking at data.
We can use our interactive viewer (Section 8.4). The following lines show how we
modified that script for the Nankai data.
8 #================================================
9 # USER AREA -- SUPPLY VALUES
10
11 # input seismic data
12 indata=Ndata/Nshots.su # SU format
13
14 # plot choices
15 myperc=95 # perc value for plot
16 plottype=0 # 0 = wiggle plot, 1 = image plot
17 Wplot=500 # Width of plot (pixels)
18 Hplot=700 # Height of plot (pixels)
19
20 # processing variables
21 sortkey=fldr # sort key (usually fldr or cdp)
22 firsts=1687 # first sort (fldr or cdp) value
23 lasts=2012 # last sort (fldr or cdp) value
24 increment=20 # sort key increment
25 tracekey=offset # trace label key
26
27 #================================================
One of the important reasons we examine the shot gathers is to determine whether we
have bad hydrophones (or for land data, geophones). A bad “phone” is easy to find –
usually the entire trace has extremely high amplitudes or zero amplitudes (a “dead”
trace). An extremely high amplitude (noisy) trace should usually have its amplitudes
replaced by zeros (“killed”); a dead trace usually does not cause processing problems.
If you use the parameters shown above, you will see the images shown in the upper
part of Figure 10.4. You might also have to wait a little while for those plots – each time
a gather is selected, you have to wait for suwind to scroll through the entire 423 Mbytes.
An alternative QC procedure is to use a different input file, a time-windowed version of
Nshots.su. We used the following command:
suwind < Nshots.su tmin=5.8 tmax=7.8 > NshotsW.su
The output file, NshotsW.su, occupies only 81 Mbytes. The viewer operates much faster
with this smaller data set. (We also changed viewer parameter Hplot to 500.)

10-5
 Nankai Data: Examine, Resample, Sort, Gain
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 10.4: Upper left: Gather 1707, 0-11 sec (30 traces). Upper right: Gather 1847,
0-11 sec (69 traces). Lower left: Gather 1707, 5.8-7.8 sec. Lower right: Gather 1847,
5.8-7.8 sec.
This data set does not appear to have any noisy traces. (The original processors
probably deleted or killed noisy traces. The left plots in Figure 10.4 show a killed trace!)

10-6
 Nankai Data: Examine, Resample, Sort, Gain

Note: While the smaller data set is faster to view, you do not see the entire gather.
Keep in mind that it is the processor’s responsibility to see and comprehend the data. The
first time we viewed the Nankai shot gathers, we viewed every 10th gather using all 11
seconds. A problem found early saves much time later!
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

10.4 f-x Display, Filter, Resample, and Sort Nankai Data

The last time we sorted shot gathers to CMP gathers (Section 7.2), we used script
sort2cmp.sh to do two tasks:
1. In the file of shot gathers, create header cdp and assign values to it.
2. Sort the file of shot gathers to CMP order.
We do not need to do the first task because, as Table 10.2 shows, Nshots.su already has
cdp values. If we only want to do the second task, we would use the command below:
susort < Nshots.su > Ncdps.su cdp offset
However, we already decided (Section 10.1.2) to change the data sample interval from 2
ms to 4 ms. Below is the documentation for suresamp:
sudoc suresamp
1 SURESAMP - Resample in time
2
3 suresamp <stdin >stdout [optional parameters]
4
5 Required parameters:
6 none
7
8 Optional Parameters:
9 nt=tr.ns number of time samples on output
10 dt=tr.dt/10^6 time sampling interval on output
11 tmin=tr.delrt/10^3 first time sample on output
12
13 Example 1: (assume original data had dt=.004 nt=256)
14 sufilter <data f=40,50 amps=1.,0. |
15 suresamp nt=128 dt=.008 | ...
16 Note the typical anti-alias filtering before sub-sampling.
17 Example 2: (assume original data had dt=.004 nt=256)
18 suresamp <data nt=512 dt=.002 | ...
19
20 Credits:
21 CWP: Dave (resamp algorithm), Jack (SU adaptation)
22
23 Trace header fields accessed: ns, dt, delrt
24 Trace header fields modified: ns, dt, delrt(only, when set tmin)
25
(Notice that instead of simply entering the name of the SU program, we entered
“sudoc” and the name of the program. With this, we find that we occasionally unearth a
valuable bit of information.)
We placed line numbers on the left for discussion; they are not part of the
documentation. Line 16 suggests that we use sufilter before we resample the data. The
example filter, line 14, begins the high-cut at 40 Hz and by 50 Hz frequencies have zero
amplitude.
We accept the recommendation to use an anti-alias filter, a high-cut filter. The alias
frequency for 2 ms data is 250 Hz (1/(2*sample interval)) and the alias
frequency for 4 ms data is 125 Hz. (Two other names for the alias frequency are the

10-7
 Nankai Data: Examine, Resample, Sort, Gain

Nyquist frequency and the folding frequency. References: “Frequency Aliasing,” Yilmaz,
1987, p. 11 and Yilmaz, 2001, p. 30.) Before we resample our 2 ms data, we need to
eliminate frequencies higher than 125 Hz to avoid “folding” them into lower frequencies
during resampling.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

10.4.1 Script fxdisp

The suresamp filter example might be too drastic for our data. That is, we might not
want to remove all the frequencies from 50 Hz to the Nyquist frequency of 125 Hz. We
can use script fxdisp.sh to see a gather and its frequencies.
The fxdisp.sh dialog asks for three values: the input file name, a perc value for the
wiggle plot of the input file, and a key (trace header) to label the x-axis of the wiggle
plot. The script creates two displays: a wiggle plot of the input traces (lines 40-44) and a
frequency display (lines 33-37). The script transforms each input seismic trace to a
frequency trace (line 30); the frequency traces are unrelated to each other. The instruction
on line 30
suop op=norm
divides each trace by its maximum value. Since this instruction balances each trace,
suximage (lines 33-37) does not use the perc parameter; it is unnecessary.
The script repeats these three questions and two displays until you quit the script. No
data are saved after processing.
If you make a mistake while typing, use the Delete key, not the Back Space key.
1 #! /bin/sh
2 # File: fxdisp.sh
3 # Start this script with fxdisp.scr
4 #
5 # Transform each input trace to its frequency spectrum
6 #
7 # Inputs (2): gather name, perc, x-axis key
8 # Outputs (2): f-x display, gather display
9 #
10 # This is only a display script. Data are not saved.
11
12 echo " "
13 echo " *** F-X SPECTRAL DISPLAY ***"
14 echo " "
15
16 #------------------------------------------------
17 # Begin loop
18 #------------------------------------------------
19
20 ok=false
21 while [ $ok = false ]
22 do
23
24 echo " Enter file name perc value key for x-axis"
25 echo " Example: dog.su 90 offset"
26 > /dev/tty
27 read indata myperc xkey
28
29 # Create frequency spectrum of each trace and normalize
30 suspecfx < $indata | suop op=norm |
31
32 # Plot frequency traces
33 suximage xbox=10 ybox=10 wbox=400 hbox=600 \

10-8
 Nankai Data: Examine, Resample, Sort, Gain

34 label1=" Freq (Hz)" \

35 windowtitle="f-x spectrum" title=$indata \
36 cmap=hsv7 legend=1 unit=Amplitude \
37 verbose=0 bclip=0.5 wclip=0.0 &
38
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

39 # Plot input seismic data

40 suxwigb < $indata perc=$myperc key=$xkey \
41 xbox=420 ybox=10 wbox=400 hbox=600 \
42 label1=" Time (s)" label2=$xkey \
43 windowtitle="t-x data" title=$indata \
44 verbose=0 &
45
46 #------------------------------------------------
47 # Choose loop or exit
48 #------------------------------------------------
49
50 echo " "
51 echo " 1 = again 2 = exit"
52 > /dev/tty
53 read selection
54 case $selection in
55 1)
56 ok=false
57 ;;
58 2)
59 ok=true
60 zap xwigb > tmp1 # decrease screen messages
61 zap ximage > tmp1 # decrease screen messages
62 ;;
63 esac
64 done
65
66 #------------------------------------------------
67 # Exit
68 #------------------------------------------------
69
70 rm -f tmp1
71 exit
72
We use script fxdisp.scr to run fxdisp.sh:
1 #! /bin/sh
2 # File: fxdisp.scr
3 # Run this script to start script fxdisp.sh.
4 #
5 # This is only a display script. Data are not saved.
6
7 xterm -geom 80x20+10+640 -e fxdisp.sh
8
We will not run all the shot gathers through fxdisp; we will window two shot gathers
from the 2-D line with the following commands:
suwind < Ndata/Nshots.su key=fldr min=1707 max=1707 > Nshot1707.su
suwind < Ndata/Nshots.su key=fldr min=1847 max=1847 > Nshot1847.su
Figure 10.5 shows the fxdisp plots. Typical seismic data has good high frequency
content up to around 50-70 Hz. Frequencies above this are usually noise (unless the
survey was designed specifically to record high-frequencies – frequencies up to around
110-150 Hz). Based on these plots, we will design the anti-alias filter to start at 85 Hz
and completely zero frequencies at 95 Hz, safely below the 4 ms Nyquist frequency of
125 Hz.

10-9
 Nankai Data: Examine, Resample, Sort, Gain

Remember that if you click the middle mouse button in either plot, information of
where you click is shown in the upper left corner of the plot. A click in the left plot shows
frequency-trace mouse location; a click in the right plot shows time-trace mouse location.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 10.5: Top: Shot gather 1707, f-x plot and x-t plot. Bottom: Shot gather 1847, f-x
plot and x-t plot.
Also, based on examining the f-x plots of Figure 10.5, we will remove low
frequencies that appear to be random signal. We will begin the low-frequency taper at 16
Hz and completely pass frequencies at 21 Hz
The x-axis of the f-x plot is labeled with tracr values (the default key); this file does
not have valid f2 key values (see Section 7.7).

10-10
 Nankai Data: Examine, Resample, Sort, Gain

10.4.2 Anti-alias Filter, Resample, and Sort

Script, Nsort2cmp.sh, filters, resamples, and sorts the shot gathers to CMP gathers.
1 #! /bin/sh
2 # File: Nsort2cmp.sh
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

3
4 # - Filter: cut freq below 21 Hz & above 95 Hz (anti-alias)
5 # - Resample shot gathers from 2 ms to 4 ms sample interval
6 # - Sort to cdp order
7
8 indata=Ndata/Nshots.su
9 outdata=Ndata/Ncdps4.su
10
11 sufilter < $indata f=16,21,85,95 amps=0,1,1,0 |
12 suresamp nt=2750 dt=.004 |
13 susort > $outdata cdp offset
14
15 # Exit politely from shell
16 exit
17
Note: The size of Ncdps4.su is 214 Mbytes. If you work in a group, we suggest that
only one person sort the shot gathers and everyone in the group use a soft link (Section
10.2) to refer to the file of CMP gathers.
Below is the surange output of Ncdps4.su.
surange < Ncdps4.su
19057 traces:
tracl=(1,19057) tracr=(1,19057) fldr=(1687,2012) tracf=(28,96) cdp=(900,1300)
cdpt=(1,69) trid=(1,2) offset=(-2435,-170) scalel=-10000 scalco=-10000
counit=1 muts=(0,11000) ns=2750 dt=4000 day=206
hour=(21,22) minute=(0,59) sec=(1,59)

Figure 10.6: Fold values for data set Ncdps4.su.

10-11
 Nankai Data: Examine, Resample, Sort, Gain

We have 401 gathers. But, the first gathers of the line are not full fold (Figure 10.6).
(We describe the program, sukeycount, which generated the data for Figure 10.6 in
Appendix C.) Upon inspection, we find that the first full-fold CMP gather (48 traces) is
CMP 933.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

To make reasonably detailed velocity analysis at a regular interval along the line, we
choose to analyze every 25th CMP starting with CMP 933; that is, fifteen full-fold CMPs:
933, 958, 983, 1008, … 1208, 1233, 1258, 1283.

Figure 10.7: Velocity analysis locations (CMPs) based on Figure 10.3, right.

10-12
 Nankai Data: Examine, Resample, Sort, Gain

10.5 Gain -- Spherical Divergence Correction

To apply spherical divergence correction to the Nankai data, we could simply use “2”
in the tpow option of sugain:
sugain < Ncdps4.su tpow=2 > Ncdps4g.su
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

We might do this because we know that, theoretically, a spherical wavefront weakens as

a function of 1/distance squared. However, real data are not always as theory suggests.
To determine which exponential gain to apply to the entire line of CMPs, we will
apply gain to two CMPs: 1055 and 1280. We window them from the 2-D line:
suwind < Ndata/Ncdps4.su key=cdp min=1055 max=1055 > Ncdp1055.su
suwind < Ndata/Ncdps4.su key=cdp min=1280 max=1280 > Ncdp1280.su
Interactive gain script, igain, allows us to use most of the options in sugain. We can
also apply more than one gain to the same file by re-using the gained file.
Script igain.scr can be used to run igain.sh:
1 #! /bin/sh
2 # File: igain.scr
3 # Run this script to start script igain.sh
4
5 xterm -geom 80x15+10+640 -e igain.sh
6
Before running igain.sh we supply the name of the seismic file on line 11 and a perc
value on line 12.
The script creates three displays for the original data and after every time gain is
applied (Figures 10.8 and 10.9):
• a wiggle plot of the file,
• a decibel (dB) ximage plot of the user-selected trace(s), and
• an amplitude graph of the user-selected trace(s).
1 #! /bin/sh
2 #File: igain.sh
3 # Run script igain.scr to start this script
4
5 # Set messages on
6 ##set -x
7
8 #================================================
9 # User-supplied values
10
11 indata=Ncdp1280.su # Input file
12 myperc=95 # perc value
13
14 #================================================
15
16 echo " "
17 echo " *** INTERACTIVE GAIN TEST ***"
18 echo " "
19
20 # Remove temporary files
21 rm -f tmp*
22
23 #------------------------------------------------
24 # Get preliminary information: trace(s) to analyze
25 #------------------------------------------------
26
27 # Ask user for display key

10-13
 Nankai Data: Examine, Resample, Sort, Gain

28 echo " "

29 echo " Supply name of key for trace selection."
30 echo " For example: offset or tracr"
31 > /dev/tty
32 read mykey
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

33
34 echo " "
35 echo " Supply minimum and maximum key values for dB amplitude display."
36 echo " For the best display, use only one to three traces."
37 echo " for example: 300 300"
38 echo " or: 250 450"
39 echo " or: -450 -250"
40 > /dev/tty
41 read mykey1 mykey2
42
43 #------------------------------------------------
44 # Log preliminary information
45
46 echo " *** INTERACTIVE GAIN TEST ***" > tmp4
47 echo "Input file = $indata perc = $myperc" >> tmp4
48 echo "key = $mykey min value = $mykey1 max value = $mykey2" >> tmp4
49
50 #------------------------------------------------
51 # Show original gather and spectra first
52 #------------------------------------------------
53
54 suxwigb < $indata xbox=10 ybox=10 wbox=400 hbox=600 \
55 label1=" Time (s)" label2="$mykey" \
56 title="Original gather" key=$mykey \
57 perc=$myperc verbose=0 &
58
59 suwind < $indata key=$mykey min=$mykey1 max=$mykey2 |
60 suattributes mode=amp |
61 suop op=db > tmp0
62
63 suximage < tmp0 xbox=420 ybox=10 wbox=190 hbox=600 \
64 label1=" Time" label2="Amplitude" title="Amplitude" \
65 grid1=dot grid2=dot legend=1 units=dB \
66 cmap=hsv1 verbose=0 &
67
68 suxgraph < tmp0 -geometry 190x600+620+10 \
69 label1="Time" label2="Amplitude" \
70 title="$mykey $mykey1 $mykey2" grid1=dot grid2=dot \
71 nTic2=2 -bg white verbose=0 &
72
73 #------------------------------------------------
74 # Amplitude correction
75 #------------------------------------------------
76
77 new=true # true = first test
78 ok=false # false = continue looping
79
80 while [ $ok = false ]
81 do
82
83 rm -f tmp0 # remove earlier copy of file to be gained
84
85 if [ $new = true ] ; then
86 cp $indata tmp0
87 echo " -> Original data" >> tmp4
88 else
89 echo " "
90 echo "Enter A to add another gain correction"

10-14
 Nankai Data: Examine, Resample, Sort, Gain

91 echo "Enter S to start over"

92 > /dev/tty
93 read choice1
94
95 case $choice1 in
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

96 [sS])
97 cp $indata tmp0
98 echo " -> Using original data"
99 echo " -> Using original data" >> tmp4
100 ;;
101 [aA])
102 cp tmp1 tmp0
103 echo " -> Using modified data"
104 echo " -> Using modified data" >> tmp4
105 ;;
106 esac
107
108 fi
109
110 echo " "
111 echo "Select Gain Correction Method:"
112 echo " Enter A for automatic gain correction"
113 echo " Enter B to add an overall bias value"
114 echo " Enter C to clip data"
115 echo " Enter E to multiply data by exp(t*epow)"
116 echo " Enter J to use Jon Claerbout values:"
117 echo " tpow=2 gpow=.5 qclip=.95"
118 echo " Enter M to balance by dividing by mean"
119 echo " Enter R to balance data by 1/rms"
120 echo " Enter S to scale data"
121 echo " Enter T to multiply data by t^tpow"
122 > /dev/tty
123 read choice2
124
125 case $choice2 in
126 [aA])
127 echo " Supply window length in seconds:"
128 > /dev/tty
129 read wagc
130 echo " -> AGC: window length = $wagc s"
131 echo " -> AGC: window length = $wagc s" >> tmp4
132 sugain < tmp0 agc=1 wagc=$wagc > tmp1
133 ;;
134 [bB])
135 echo " Supply overall bias value"
136 > /dev/tty
137 read bias
138 echo " -> Bias with value $bias"
139 echo " -> Bias with value $bias" >> tmp4
140 sugain < tmp0 bias=$bias > tmp1
141 ;;
142 [cC])
143 echo " Supply clip value between 0.00 and 1.00:"
144 > /dev/tty
145 read qclip
146 echo " -> Clip by $qclip of absolute values"
147 echo " -> Clip by $qclip of absolute values" >> tmp4
148 sugain < tmp0 qclip=$qclip > tmp1
149 ;;
150 [eE])
151 echo " Supply exponent epow:"
152 > /dev/tty
153 read epow

10-15
 Nankai Data: Examine, Resample, Sort, Gain

154 echo " -> Gain function is: A'=A*e^(t*$epow)"

155 echo " -> Gain function is: A'=A*e^(t*$epow)" >> tmp4
156 sugain < tmp0 epow=$epow > tmp1
157 ;;
158 [jJ])
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

159 echo " -> Jon: tpow=2 gpow=.5 qclip=.95"

160 echo " -> Jon: tpow=2 gpow=.5 qclip=.95" >> tmp4
161 sugain < tmp0 jon=1 > tmp1
162 ;;
163 [mM])
164 echo " -> Balance: divide by mean"
165 echo " -> Balance: divide by mean" >> tmp4
166 sugain < tmp0 mbal=1 > tmp1
167 ;;
168 [rR])
169 echo " -> Balance: divide by rms value"
170 echo " -> Balance: divide by rms value" >> tmp4
171 sugain < tmp0 pbal=1 > tmp1
172 ;;
173 [sS])
174 echo " Supply overall scale factor"
175 > /dev/tty
176 read scale
177 echo " -> Scale with factor $scale"
178 echo " -> Scale with factor $scale" >> tmp4
179 sugain < tmp0 scale=$scale > tmp1
180 ;;
181 [tT])
182 echo " Supply exponent tpow:"
183 > /dev/tty
184 read tpow
185 echo " -> Gain function is: A'=A*t^$tpow"
186 echo " -> Gain function is: A'=A*t^$tpow" >> tmp4
187 sugain < tmp0 tpow=$tpow > tmp1
188 ;;
189 esac
190
191 #------------------------------------------------
192 # Plot gained data
193 #------------------------------------------------
194
195 suxwigb < tmp1 xbox=420 ybox=10 wbox=400 hbox=600 \
196 label1=" Time (s)" label2="$mykey" \
197 title="Gain applied" key=$mykey \
198 perc=$myperc verbose=0 &
199
200 suwind < tmp1 key=$mykey min=$mykey1 max=$mykey2 |
201 suattributes mode=amp |
202 suop op=db > tmp2
203
204 suximage < tmp2 xbox=830 ybox=10 wbox=190 hbox=600 \
205 label1=" Time" label2="Amplitude" title="Amplitude" \
206 grid1=dot grid2=dot legend=1 units=dB \
207 cmap=hsv1 verbose=0 &
208
209 suxgraph < tmp2 -geometry 190x600+1030+10 \
210 label1="Time" label2="Amplitude" \
211 title="$mykey $mykey1 $mykey2" grid1=dot grid2=dot \
212 nTic2=2 -bg white verbose=0 &
213
214 #------------------------------------------------
215 # Choose loop or exit
216 #------------------------------------------------

10-16
 Nankai Data: Examine, Resample, Sort, Gain

217
218 echo " "
219 echo "Enter 1 for more Amplitude corrections"
220 echo "Enter 2 to output gained seismic data and EXIT"
221 > /dev/tty
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

222 read choice3

223
224 case $choice3 in
225 1)
226 ok=false
227 ;;
228 2)
229 cp tmp1 igain.su
230 echo " *** Output data file: igain.su"
231 echo " *** Output data file: igain.su" >> tmp4
232 cp tmp4 igain.txt
233 echo " *** Output log file: igain.txt"
234 pause exit
235 ok=true
236 zap xwigb > tmp3 # decrease screen messages
237 zap ximage > tmp3 # decrease screen messages
238 zap xgraph > tmp3 # decrease screen messages
239 ;;
240 esac
241
242 new=false # true = first test
243
244 done
245
246 #------------------------------------------------
247 # Exit
248 #------------------------------------------------
249
250 # Remove temporary files
251 rm -f tmp*
252
253 # Exit politely from shell
254 exit
255
The script applies gain to the entire file, but only the user-selected trace or traces are
used to create the second and third displays. In lines 27-41, the user is asked to supply the
name of a key (offset, tracr, etc.) and corresponding key minimum and key maximum
values. We selected key offset and supplied -1100 -1000 for our offsets. We selected the
offsets by plotting the CMP and zooming with the left mouse (Section 1.10).
suxwigb < Ncdp1280.su perc=95 key=offset
We also used surange and sugethw to examine trace headers:
sugethw < Ncdp1280.su key=tracr,offset,tracf > headers1280.txt
When we looked at our output of sugethw, file headers1280.txt, we saw that there are
two offsets in this range: -1036 and -1003. Two traces are appropriate for this script. We
selected offsets away from the direct arrival and that have good reflection character. If,
instead, we had selected tracr and values 18081 18082, the x-axis values of the wiggle
plot would have matched the x-axis values of the dB (second) display (Section 7.7).
Lines 54-71 display the input data. Line 59 windows the selected traces for the
second and third display. Line 60 creates the trace attribute of instantaneous amplitude
(amplitude envelope traces) and line 61 converts the amplitudes to dB.

10-17
 Nankai Data: Examine, Resample, Sort, Gain

We applied (t^pow) gain, (option T), three times, each time independent of the others.
(See the log file below.) Table 10.4 shows how the various gain selections enhanced the
data’s dynamic range.
Table 10.4: dB ranges of Figures 10.8 and 10.9
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

No gain tpow = 1.8 tpow = 2.2 tpow = 2.5

maximum dB 0 0 0 0
minimum dB -66 -135.6 -157.1 -173.5
Lines 111-189 are where gain is selected and applied. If the user chooses to apply
gain again, the user can use the original file (indata) or the data that have been gained.
Lines 90-106 monitor the choice. The data to be gained are in tmp0 and after gain the
data are in tmp1. If the user chooses a fresh gain experiment, line 97 copies the input
data, indata, to tmp0; if the user chooses to gain previously gained data, line 102 copies
the gained data, tmp1, to tmp0.
The script writes the final gained data to disk (line 229 creates file igain.su) and a text
(ASCII) file of the processing choices (line 232 creates file igain.txt). We suggest that
you rename the two files to avoid overwriting them by a later run.
Below are the contents of igain.txt:
*** INTERACTIVE GAIN TEST ***
Input file = Ncdp1280.su perc = 95
key = offset min value = -1100 max value = -1000
-> Original data
-> Gain function is: A'=A*t^1.8
-> Using original data
-> Gain function is: A'=A*t^2.2
-> Using original data
-> Gain function is: A'=A*t^2.5
*** Output data file: igain.su
To select a tpow value to apply to the entire line, we principally rely upon an
examination of the third display. We ignore information above six seconds, because that
time is when the acoustic energy is in the water layer. Below six seconds, we are looking
for a vertical amplitude line. In all three cases (tpow = 1.8, 2.2, 2.5), the line is close to
vertical. In the third trial, tpow = 2.5, the amplitude line is beginning to deviate to the
right. After discussion, we decide to use 2.1; it treats the data well and is geophysically
reasonable. (CMP 1055 gave similar results.)
Finally, we apply spherical divergence correction with the command that we showed
at the beginning of this section, with our new tpow value.
sugain < Ncdps4.su tpow=2.1 > Ncdps4g.su

10-18
 Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

10-19
Nankai Data: Examine, Resample, Sort, Gain

Figure 10.8: Top: CMP 1280 without gain. Bottom: CMP 1280 with gain, tpow=1.8.
 Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

tpow=2.5.

10-20
Nankai Data: Examine, Resample, Sort, Gain

Figure 10.9: Top: CMP 1280 with gain, tpow=2.2. Bottom: CMP 1280 with gain,
 Nankai Data: Examine, Resample, Sort, Gain

10.6 Single-trace versus Multi-trace Processes

When we consider applying a process, such as spherical divergence correction,
should we apply it to the shot gathers or the CMP gathers? It might not matter. If the data
order does not matter and we are considering a single-trace process, the result is the same
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

– a trace is processed without regard to its neighbors. Some example SU single-trace

processes are listed below.
On the other hand, when we think of applying a multi-trace process, such as NMO,
the order of the data is important. Changing the order of the data (shot gathers versus
CMP gathers versus a stack file) dramatically changes the output. Some examples of SU
multi-trace processes are listed below.
Table 10.5: Compare processes
Single-trace Multi-trace
sugain – gain sunmo – NMO
sufilter – 1-D filter suvelan – semblance
sukill – kill trace sudipfilt – 2-D filter
supef – deconvolution sustolt – Stolt migration
A further question is, why are we thinking of applying a process? Spherical
divergence correction is a scientifically valid way to balance a trace in time. We need this
to re-balance data that are naturally attenuated in time. Therefore, we apply this gain
early in the processing stream.
Normal moveout correction, NMO, is meaningful to CMP gathers, but not shot
gathers. On the other hand, we might apply deconvolution, a single-trace process, twice.
The first time might be a gap deconvolution to remove multiples; the second application
would be a spiking deconvolution to sharpen the reflections (compress the wavelets).

10.7 Summary
This chapter saw us working with large files and real data. When we processed model
data, we went easily from shot gathers to sort to CMP gathers to velocity analysis. Real
data require inspection, gain, frequency analysis, and much more that we will see in later
chapters.

10-21
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

This page has been intentionally left blank

 Nankai: Velocity Analysis, NMO, Stack, Migration

11. Nankai: Velocity Analysis, NMO, Stack, Migration

11.1 Velocity Analysis
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Below is the user area of iva.sh changed for the Nankai data. As stated in Section
10.4.2, we will do velocity analysis on every 25th CMP starting with CMP 933 (lines 13-
16). We change the perc value to 95 (line 29), an accommodation for real data.
Semblance values go from 1000 m/s (line 38) to 7000 m/s (nvs*dvs + fvs). CVS panels
range from 1000 m/s to 6000 m/s (lines 41-44).
8 #================================================
9 # USER AREA -- SUPPLY VALUES
10 #------------------------------------------------
11 # CMPs for analysis
12
13 cmp1=933 cmp2=958 cmp3=983 cmp4=1008
14 cmp5=1033 cmp6=1058 cmp7=1083 cmp8=1108
15 cmp9=1133 cmp10=1158 cmp11=1183 cmp12=1208
16 cmp13=1233 cmp14=1258 cmp15=1283
17
18 numCMPs=15
19
20 #------------------------------------------------
21 # File names
22
23 indata=Ndata/Ncdps4g.su # SU format
24 outpicks=Nvpick.txt # ASCII file
25
26 #------------------------------------------------
27 # display choices
28
29 myperc=95 # perc value for plot
30 plottype=1 # 0 = wiggle plot, 1 = image plot
31
32 #------------------------------------------------
33 # Processing variables
34
35 # Semblance variables
36 nvs=240 # number of velocities
37 dvs=25 # velocity intervals
38 fvs=1000 # first velocity
39
40 # CVS variables
41 fc=1000 # first CVS velocity
42 lc=6000 # last CVS velocity
43 nc=10 # number of CVS velocities (panels)
44 XX=11 # ODD number of CMPs to stack into central CVS
45
46 #================================================
By looking at the Nankai stack, Figure 10.3 left, we know the water layer is
approximately 6 seconds (two-way time) and we think there is not much reflection
character below 7.8 seconds. The left side of Figure 11.1 is the velan of CMP 933. It
appears that the time for picking events is restricted to between 6 seconds and 8 seconds.

11-1
 Nankai: Velocity Analysis, NMO, Stack, Migration
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 11.1: Left: Velan of CMP 933, 0-11 seconds. Right: Velan of CMP 933, 5.5-11
seconds.
Instead of making velocity picks in a narrow space on the screen, we window the file
of CMPs:
suwind < Ncdps4g.su tmin=5.5 tmax=11 > Ncdps4g5511.su
and use this as the input to iva.sh
23 indata=Ndata/Ncdps4g5511.su # SU format
We chose the full time (tmax=11) because on CMP 933 we see events around 9.5
seconds and we don’t want to miss these and other potentially later events.
The result of using this new input file is shown on the right side of Figure 11.1. This
makes better use of the screen.
Because we are not picking from zero time, we will have to modify the output pick
file, adding zero time to the beginning of all the tnmo series and adding the water velocity
to the beginning of all the vnmo series.
Below is the output file, Nvpick.txt. For the sake of presentation, we made line 1 into
two lines. Line 1 as presented here has no spaces at the end of the first part and no spaces
at the beginning of the second part, making this a usable file.
1 cdp=933,958,983,1008,1033,1058,1083,\
1108,1133,1158,1183,1208,1233,1258,1283 \
2 #=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15 \
3 tnmo=5.52263,6.44307,6.97119,7.65775,8.50274,10.9849 \
4 vnmo=1522.81,1460.57,1622.4,1834.01,2331.93,2892.08 \
5 tnmo=5.53018,6.49588,7.55967,9.19685,10.9774 \
6 vnmo=1497.92,1497.92,1634.84,2966.77,3016.56 \
7 tnmo=5.52263,6.35254,7.40878,9.27229,10.9849 \

11-2
 Nankai: Velocity Analysis, NMO, Stack, Migration

8 vnmo=1497.92,1460.57,1659.74,2307.03,2991.67 \
9 tnmo=5.53772,7.48422,8.93279,10.9925 \
10 vnmo=1497.92,1634.84,2282.14,3564.27 \
11 tnmo=5.55281,6.99383,7.5144,8.48011,10.9925 \
12 vnmo=1522.81,1659.74,1734.43,2319.48,2991.67 \
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

13 tnmo=5.53018,6.69959,7.46159,9.45336,10.9774 \
14 vnmo=1485.47,1572.6,1871.35,2282.14,2929.43 \
15 tnmo=5.53018,6.79012,7.58985,8.88752,10.9774 \
16 vnmo=1497.92,1622.4,1771.77,2406.61,2979.22 \
17 tnmo=5.53772,6.29973,7.3786,8.7668,10.9698 \
18 vnmo=1497.92,1497.92,1759.32,2842.29,3589.17 \
19 tnmo=5.53772,7.30315,9.16667,10.9925 \
20 vnmo=1510.36,1672.19,2630.68,3053.91 \
21 tnmo=5.52263,7.39369,9.60425,10.9849 \
22 vnmo=1497.92,1721.98,2443.96,3004.11 \
23 tnmo=5.53018,6.93347,7.55213,10.9849 \
24 vnmo=1535.26,1634.84,1734.43,2867.19 \
25 tnmo=5.52263,7.43896,8.2915,10.9849 \
26 vnmo=1510.36,1647.29,2282.14,2979.22 \
27 tnmo=5.53018,7.49931,8.13306,10.9698 \
28 vnmo=1510.36,1659.74,1921.15,2555.99 \
29 tnmo=5.54527,7.35597,10.9849 \
30 vnmo=1510.36,1697.08,3004.11 \
31 tnmo=5.53772,7.29561,8.19342,9.08368,10.9925 \
32 vnmo=1522.81,1746.88,2456.41,2879.64,4248.91 \
Our picking philosophy was simple: slowly increasing velocity with depth. We did
not consider any picks that deviated strongly from this. We think this is adequate for a
first look at stack and migration. (In the beginning, simple is good.) After we examine the
stack data and the migrated data, we can reconsider our picks.
The file Nvpick.txt must be modified before it can be used for NMO:
• Line 2 must be removed because a comment line cannot be in the midst of a
command. The continuation mark “\” continues a command onto the next line.
(However, a comment line can be between piped (|) commands.)
• We picked from a seismic file that did not go to zero time, so we must add the
zero time and the velocity at zero time to the tnmo and vnmo lines. The result is in
the script in the next section.
Note: If we had windowed the input file to exclude the last time; for example, if we
had started the window at 5.5 seconds and ended the window at 10 seconds
suwind < Ncdps4g.su tmin=5.5 tmax=10 > Ncdps4g5510.su
we would have to edit Nvpick.txt at the start of each series and at the end of each series.
At the end of each tnmo series, we would add 11.0 and at the end of each vnmo series we
would repeat the last picked velocity.
Because of the rough sea-floor topography and the spatial-temporal interpolation of
the velocity profiles, it may be wise to pick velocities at much more closely spaced CMP
locations. We leave this exercise to the reader who should be able to improve on our final
image by performing this additional task.

11.2 NMO and Stack

File Nnmo.sh is applies NMO and stack to the 2-D line of gained CMPs. The size of
the stack file is only 4.5 Mbytes. Notice that the tnmo lines have 0.0 as the first value and
the vnmo lines have 1500. as the first value.

11-3
 Nankai: Velocity Analysis, NMO, Stack, Migration

For the sake of presentation, we made line 18 into two lines. Line 18 as presented
here has no spaces at the end of the first part and no spaces at the beginning of the second
part, making this a usable file.
1 #! /bin/sh
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

2 # File: Nnmo.sh
3 # Apply NMO (flatten) 2-D line of CMPs
4 # Input (1): 2-D line of CMPs
5 # Output (1): NMO-corrected 2-D line of CMPs
6 # Use: Nnmo.sh
7 #
8 # NMO correction is interpolated between named CMPs.
9
10 # Set debugging on
11 set -x
12
13 # Name data sets
14 indata=Ndata/Ncdps4g.su
15 outdata=Nstack4.su
16
17 sunmo < $indata \
18 cdp=933,958,983,1008,1033,1058,1083,\
1108,1133,1158,1183,1208,1233,1258,1283 \
19 tnmo=0.0,5.52263,6.44307,6.97119,7.65775,8.50274,10.9849 \
20 vnmo=1500.,1522.81,1460.57,1622.4,1834.01,2331.93,2892.08 \
21 tnmo=0.0,5.53018,6.49588,7.55967,9.19685,10.9774 \
22 vnmo=1500.,1497.92,1497.92,1634.84,2966.77,3016.56 \
23 tnmo=0.0,5.52263,6.35254,7.40878,9.27229,10.9849 \
24 vnmo=1500.,1497.92,1460.57,1659.74,2307.03,2991.67 \
25 tnmo=0.0,5.53772,7.48422,8.93279,10.9925 \
26 vnmo=1500.,1497.92,1634.84,2282.14,3564.27 \
27 tnmo=0.0,5.55281,6.99383,7.5144,8.48011,10.9925 \
28 vnmo=1500.,1522.81,1659.74,1734.43,2319.48,2991.67 \
29 tnmo=0.0,5.53018,6.69959,7.46159,9.45336,10.9774 \
30 vnmo=1500.,1485.47,1572.6,1871.35,2282.14,2929.43 \
31 tnmo=0.0,5.53018,6.79012,7.58985,8.88752,10.9774 \
32 vnmo=1500.,1497.92,1622.4,1771.77,2406.61,2979.22 \
33 tnmo=0.0,5.53772,6.29973,7.3786,8.7668,10.9698 \
34 vnmo=1500.,1497.92,1497.92,1759.32,2842.29,3589.17 \
35 tnmo=0.0,5.53772,7.30315,9.16667,10.9925 \
36 vnmo=1500.,1510.36,1672.19,2630.68,3053.91 \
37 tnmo=0.0,5.52263,7.39369,9.60425,10.9849 \
38 vnmo=1500.,1497.92,1721.98,2443.96,3004.11 \
39 tnmo=0.0,5.53018,6.93347,7.55213,10.9849 \
40 vnmo=1500.,1535.26,1634.84,1734.43,2867.19 \
41 tnmo=0.0,5.52263,7.43896,8.2915,10.9849 \
42 vnmo=1500.,1510.36,1647.29,2282.14,2979.22 \
43 tnmo=0.0,5.53018,7.49931,8.13306,10.9698 \
44 vnmo=1500.,1510.36,1659.74,1921.15,2555.99 \
45 tnmo=0.0,5.54527,7.35597,10.9849 \
46 vnmo=1500.,1510.36,1697.08,3004.11 \
47 tnmo=0.0,5.53772,7.29561,8.19342,9.08368,10.9925 \
48 vnmo=1500.,1522.81,1746.88,2456.41,2879.64,4248.91 |
49 sustack > $outdata
50
51 # Exit politely from shell
52 exit
53
We plot the stack file with suximage.
suximage < Nstack4.su key=cdp perc=95 &

11-4
 Nankai: Velocity Analysis, NMO, Stack, Migration

We see that our stack file (Figure 11.2) does not resemble the original stack file
(Figure 10.3, right). We suspect the original file Nstack.su is actually a stacked and
migrated file. (In fact, this is the case, as is clear from the publication by Moore et al.,
1990.)
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 11.2: Stack data set Nstack4.su.

11-5
 Nankai: Velocity Analysis, NMO, Stack, Migration

11.3 Stolt Migration with sustolt

11.3.1 Script Nmigcvp.sh

Below is the user area of Nmigcvp.sh for the Nankai data. The CDP bin distance
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

(reported in Moore, et al., 1990) is 16.667 m (parameter dxcdp, line 22). Because we do
not know rms velocities for these data, we set vscale=1.0 (line25). We process for image,
not for velocities. Our first velocity is 1400 m/s and our last is 2000 m/s, a narrow range.
11 #================================================
12 # USER AREA -- SUPPLY VALUES
13 #------------------------------------------------
14
15 # Seismic files
16 indata=Nstack4.su # SU format
17 outdata=Ndata/Nmigcvp.su # migration Constant Velocity Panels
18
19 # Migration variables
20 cdpmin=900 # Start CDP value
21 cdpmax=1300 # End CDP value
22 dxcdp=16.667 # distance between adjacent CDP bins (m)
23 smig=1.0 # stretch factor (0.6 typical if vrms increasing)
24 # [the "W" factor] (Default=1.0)
25 vscale=1.0 # scale factor to apply to velocities (Default=1.0)
26 lstaper=20 # length of side tapers (traces) (Default=0)
27 lbtaper=100 # length of bottom taper (samples) (Default=0)
28
29 # Velocity panel variables
30 firstv=1400 # first velocity value
31 lastv=2000 # last velocity value
32 increment=200 # velocity increment
33
34 numVtest=100 # use to limit number of velocity panels
35 # otherwise, use very large value (100)
36
37 #================================================

11.3.2 Script Niviewcvp

Below is the user area of Niviewcvp.sh for the Nankai data. We set Wplot and Hplot
to accommodate a larger image.
11 #================================================
12 # USER AREA -- SUPPLY VALUES
13 #------------------------------------------------
14
15 # Input seismic data
16 indata=Nmigcvp.su # SU format
17
18 # Plot choices
19 myperc=95 # perc value for plot
20 plottype=1 # 0 = wiggle plot, 1 = image plot
21 Wplot=550 # Width of plot (pixels)
22 Hplot=700 # Height of plot (pixels)
23
24 #================================================
We also changed line 62. For Model 4, the line is:
62 suwind < $indata key=$sortkey min=$i max=$i > tmp1
To view the Nankai data, we also window time:
62 suwind < $indata key=$sortkey min=$i max=$i tmin=5.5 tmax=8 > tmp1

11-6
 Nankai: Velocity Analysis, NMO, Stack, Migration

We do not want to waste viewing space on the deep-water layer, and we see no
significant reflections below 7.8 seconds.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 11.3: Left: Original migrated data file. Right: Our constant-velocity Stolt
migration at 1600 m/s.
We find that our best match to the original data, and our best image, is the 1600 m/s
migration (Figure 11.3). Notice that the migration “smiles” above the water bottom have
been muted on the original data set, cosmetically improving the data set’s appearance.

11.3.3 Script Nmigmovie.sh

Below is script Nmigmovie.sh for the Nankai data. We set loop=2 (line 12) to run the
movie back and forth. Below, fframe=1300 and dframe=100 because we re-ran the
migration (Nmigcvp.sh) for velocities 1300-2000 m/s at 100 m/s increment.
1 #! /bin/sh
2 # File: Nmigmovie.sh
3 # Run a "movie" of the migration panels
4 # Plot "title" shows panel velocity
5 # Enter "xmovie" for mouse and keyboard options
6
7 set -x
8
9 indata=Ndata/Nmigcvp.su
10 perc=98
11
12 loop=2 # 1 = run panels forward continuously
13 # 2 = run panels back and forth continuously
14 # 0 = load all panels then stop
15

11-7
 Nankai: Velocity Analysis, NMO, Stack, Migration

16 n1=2750 # number of time samples

17 d1=0.004 # time sample interval
18 n2=401 # number of traces per panel
19 d2=1 # trace spacing
20
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

21 width=550 # width of window

22 height=700 # height of window
23
24 fframe=1300 # velocity of first panel for title annotation
25 dframe=100 # panel velocity increment for title annotation
26
27 suxmovie < $indata perc=$perc loop=$loop \
28 n1=$n1 d1=$d1 n2=$n2 d2=$d2 \
29 width=$width height=$height \
30 fframe=$fframe dframe=$dframe \
31 title="Velocity %g" &
32
33 exit
34
You can zoom the xmovie window the same way you zoom the other plot windows
(Section 1.10).
This movie also shows that, using a single migration velocity, 1600 m/s is a good
match for the original data. Note that this is just a small amount faster than the speed of
sound in the water layer, suggesting that vscale=1.0 is appropriate, and that the majority
of migration is due to propagation in the water layer.
We then used a script similar to migStolt.sh (see Section 9.5) to migrate and display
the data set. If we were satisfied with a single migration velocity, we might use a line
command (for an example, see Section 11.4).

11.3.4 Summary
The Nankai data shows many diffractions. As you look at the various migrations, you
can see that reflectors do not change position; this is generally flat geology. When
reflectors are under-migrated, the diffractions are not collapsed – they are “frowns.”
When reflectors are over-migrated, the diffractions change to “smiles.” The original data
set appears to be slightly over-migrated.
After we migrated Model 4 (in Chapter 9), we used CMP 110 to compare the stacking
velocity to the migration (rms) velocity. That was a valid comparison because we had
calibrated the velocity scale factor, vscale using our model velocity. However, we cannot
expect to derive geologic velocities by migrating the Nankai data because we did not do a
similar calibration.
The Nankai data set is difficult to process because the geology is under 4.5 km of
water. Because of this, the moveout that we expect to see and use for velocity analysis is
subtle. Proper velocity analysis of this data set demands patience and some prior
knowledge of the local velocities, particularly the water velocity. We encourage you to
improve upon our image.

11.4 Phase Shift Migration with sugazmig

Another migration program is sugazmig, an SU version of Jeno Gazdag's phase-shift
migration for zero-offset data. The following migration uses a single time-velocity

11-8
 Nankai: Velocity Analysis, NMO, Stack, Migration

function. (Program sumigpspi is an SU version of J. Gazdag's phase-shift plus

interpolation migration for zero-offset data. It can handle lateral velocity variations.)
Program sugazmig is easy to use – it gets most of the information that it needs from
the trace keys. Aside from the t-v pairs, the only parameter value that we supply below is
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

for dx, sugazmig’s name for the CDP bin distance. The velocities are an approximation
based on the previous Stolt migrations.
Note: For sugazmig, the velocities are INTERVAL VELOCITIES.
1 #! /bin/sh
2 # File: Ngazmig.sh
3 # Phase shift migration of Nankai stacked data.
4 # CDP spacing (dx) is 16.667 meters.
5
6 # Set messages on
7 set -x
8
9 sugazmig < Nstack4.su dx=16.6667 \
10 tmig=0,5,6,7,8,9 \
11 vmig=1500,1500,1950,2000,2050,2100 \
12 > Ngazmig.su
13
14 # Exit politely from shell
15 exit
16

Figure 11.4: Left: Original migrated data file. Right: Our migration using sugazmig.
This migration took half an hour to run on a Sun UltraSPARC III with four
processors. Figure 11.4 shows that our migration is a fair match to the original migration.

11-9
 Nankai: Velocity Analysis, NMO, Stack, Migration

Although sugazmig does some internal zero-padding, you can see minor migration
artifacts on the sides of the section at about 6 seconds. If you look at the bottom of the
time section (11 seconds), you can also see migration artifacts. However, the bottom
artifacts are unimportant since there are no strong reflections below 10 seconds. Program
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

sugazmig does not have taper parameters; in contrast to sustolt parameters lstaper and
lbtaper.
We used the following command to display the migrated file:
suwind < Ngazmig.su tmin=5.5 tmax=8 | suximage perc=95 title="Nankai stack +
sugazmig" &

11.5 Some Geologic Features

Figure 11.5 (below) shows a time-windowed portion of the original stacked and
migrated data set
Oceanic crust of the Pacific Plate lies under a veneer of sediments at the left side of
the image, and is subducted at the trench, which is filled with recent sediments (CDPs
200-600, approximately). To the right of these flat-lying sediments, the accretionary
prism is evident, the distorted sediment layer thickens to the right, and some extreme
topography results.

11.5.1 Bottom-simulating Reflector

Notice the reflector that seems to mimic the ocean-bottom topography, but is about
300 ms later. It cuts across the folded sediments, and therefore cannot be representative
of a depositional boundary. This bottom-simulating reflector (BSR) is a common feature
in some basins. It represents a reflection from a change in fluid phases; in most cases it
corresponds to the base of a layer of methane hydrate, a gas trapped in a frozen state
within an ice “foam.” Beneath the hydrate, the gas exists in a free state. Temperature (and
pressure) conditions dictate where this transition occurs. Because the sea floor is nearly a
constant-temperature boundary, the phase transition occurs at a more-or-less constant
depth (locally) beneath the sea floor.

11.5.2 Subducted Pacific Plate

Notice that the Pacific Plate oceanic crust appears to flatten at about 7.2 seconds,
rather than continue to increase “depth” to the right of the trench. This is an artifact of
velocity pull-up. The deformed sediment layer with seismic velocities of about 2000 m/s
replaces water with velocities of 1500 m/s. Thus, an increasing time between the water
bottom and the subducted crust implies increasing depth to the reflection, even though its
reflection time happens to be nearly constant.

11-10
 Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 11.5: Original migrated data file.

11-11
Nankai: Velocity Analysis, NMO, Stack, Migration
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

This page has been intentionally left blank

 Taiwan Data: Examine, Zero Traces, Re-examine

12. Taiwan Data: Examine, Zero Traces, Re-examine

12.1 Introduction
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Our third data set is another 2-D line of seismic data provided by Prof. Greg Moore of
the University of Hawaii. This “Taiwan” data set was collected near the coast of Taiwan
in 1995 (Berndt & Moore, 1999) by the University of Hawaii, San Jose State University,
and National Taiwan University. The size of our shot gather file, Tshot.su, is 411 Mbytes.
(We have only one Taiwan data set, the original shot gathers.) This data set is more
difficult to process than the Nankai data set.
Below is the surange output of Tshot.su.
surange < Tshot.su
25344 traces:
tracl=(114769,140112) tracr=(1,25344) fldr=(800,975) tracf=(1,144)
ep=(740,915)
cdp=(4027,4816) cdpt=(1,144) trid=1 nhs=1 offset=(-3663,-88)
sdepth=80000 swdep=(2120000,3540000) scalel=-10000 scalco=1 sx=(-4773100,-
3965250)
gx=(-4764600,-3600343) gy=(766,94805) counit=3 tstat=12 ns=3999
dt=4000 gain=9 afilf=160 afils=72 lcf=3
hcf=160 lcs=6 hcs=72 year=95 day=260
hour=(12,13) minute=(0,59) sec=(0,59)
Table 12.1: Some Tshot.su key values
Key Range
fldr 800 to 975
cdp 4027 to 4816
offset -3663 to -88
ns 3999
dt 4000

Figure 12.1: Traces per shot gather in data set Tshot.su.

12-1
 Taiwan Data: Examine, Zero Traces, Re-examine

Keys tracl and tracr number the 25344 traces (although, they use different starting
numbers).
Key fldr tells us there are 176 shot gathers (975-800+1).
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Since the number of samples per trace (ns) is 3999 and the trace sample interval (dt)
is 4000 microseconds, the trace length is 16 seconds (15.996 seconds)
Notice that the Taiwan shot gather trace headers already contain cdp values.
This data set has many acquisition key values such as source x-coordinate (sx),
receiver x-coordinate (gx), receiver y-coordinate (gy), and source depth (sdepth).
Figure 12.1 is a chart of the number of traces per shot gather. (We describe the
program, sukeycount, which generated the data for the figure in Appendix C.) The chart
shows that every shot gather has 144 traces.
We do not want to process 16 seconds of data. Let’s use the first 5 seconds:
suwind < Tshot.su tmax=5 > Tshot5.su
The size of our windowed shot gather file is 133 Mbytes.

12.2 Reminder: Symbolic Link

As we discussed in Section 10.2 for the Nankai data, we think this Taiwan shot file is
too large for everyone to have his or her own copy. We suggest everyone in a group
create a symbolic link to the data directory. Suppose the Taiwan file is in directory
/home/uofhawaii/taiwan/Tdata
and I (Forel) am in directory
/home/forel/suscripts
Here, I enter command
ln –s /home/uofhawaii/taiwan/Tdata .
Remember, the period at the end represents “here.” This command makes a symbolic link
from … to …
ln –s from to
This command has the effect of placing a virtual subdirectory (Tdata) in my
directory.
To prevent accidentally deleting the data, we recommend you set permissions so the
Taiwan file is only readable. We also recommend that you store a copy of this file in
another directory or on a DVD so you can quickly replace the deleted file.

12.3 View the Line of Shot Gathers

Before we begin processing the Taiwan data, we should examine some of the shot
gathers. This is a simple but important quality control (QC) step. Nothing is more
important than actually looking at data.
We can use our interactive viewer (Section 8.4). The following lines show how we
modified that script for the Taiwan data.
8 #================================================
9 # USER AREA -- SUPPLY VALUES
10
11 # input seismic data
12 indata=Tdata/Tshot5.su # SU format
13

12-2
 Taiwan Data: Examine, Zero Traces, Re-examine

14 # plot choices
15 myperc=95 # perc value for plot
16 plottype=0 # 0 = wiggle plot, 1 = image plot
17 Wplot=900 # Width of plot (pixels)
18 Hplot=600 # Height of plot (pixels)
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

19
20 # processing variables
21 sortkey=fldr # sort key (usually fldr or cdp)
22 firsts=800 # first sort (fldr or cdp) value
23 lasts=975 # last sort (fldr or cdp) value
24 increment=5 # sort key increment
25 tracekey=tracf # trace label key
26
27 #================================================
One of the important reasons we examine the shot gathers is to determine whether we
have bad hydrophones (or for land data, geophones). A bad “phone” is easy to find –
usually the entire trace has extremely high amplitudes or zero amplitudes (a “dead”
trace). An extremely high amplitude (noisy) trace should usually have its amplitudes
replaced by zeros (“killed”); a dead trace usually does not cause processing problems.
Figure 12.2 shows shot gather (fldr) 930. If you zoom the figure, you can see that
traces (tracf) 61, 62, and 143 seem to have anomalously high amplitudes. We confirmed
this with program sudumptrace. (We describe program sudumptrace in Appendix C.)
Below is the command we entered to use sudumptrace and the last ten lines of the
screen output. We used suwind twice to extract four traces from shot (fldr) 930.
suwind < Tdata/Tshot5.su key=fldr min=930 max=930 | suwind key=tracf min=60
max=63 | sudumptrace
1242 4.968 -1.7802e-01 6.6929e+03 6.6929e+03 -1.0794e+00
1243 4.972 -9.1427e-02 6.6929e+03 6.6929e+03 -8.5523e-01
1244 4.976 -4.6168e-01 6.6929e+03 6.6929e+03 -5.7248e-01
1245 4.980 -1.3838e-01 6.6929e+03 6.6929e+03 -6.9779e-01
1246 4.984 7.6338e-01 6.6929e+03 6.6929e+03 -1.3278e+00
1247 4.988 1.2673e+00 6.6929e+03 6.6929e+03 -1.5405e+00
1248 4.992 7.0116e-01 6.6929e+03 6.6929e+03 -1.3003e+00
1249 4.996 -8.4290e-02 6.6929e+03 6.6929e+03 -1.1742e+00
1250 5.000 1.1331e-01 6.6929e+03 6.6929e+03 -1.0646e+00
1251 5.004 3.0200e-01 6.6929e+03 6.6929e+03 -6.6413e-01
The first column is time sample, the second column is time value (seconds). The next
four columns are amplitude values for traces 60-63. The amplitude values of traces 61
and 62 are constant at 6692.9 while the amplitudes of traces 60 and 63 vary at lower
amplitudes.
Below is the command we entered to use sudumptrace on the last four traces in shot
(fldr) 930. Below that are the last ten lines of the screen output.
suwind < Tdata/Tshot5.su key=fldr min=930 max=930 | suwind key=tracf min=141
max=144 | sudumptrace
1242 4.968 -3.2838e+00 -2.9376e-01 1.7135e+03 -6.8847e-01
1243 4.972 -3.4147e+00 9.3122e-01 1.7135e+03 -6.8070e-01
1244 4.976 -2.8242e+00 1.7697e+00 1.7135e+03 4.7962e-01
1245 4.980 -3.0289e+00 9.1051e-01 1.7135e+03 5.8127e-01
1246 4.984 -3.4756e+00 4.9018e-01 1.7135e+03 -8.4847e-01
1247 4.988 -3.3356e+00 1.4016e+00 1.7135e+03 -8.5439e-01
1248 4.992 -3.4664e+00 1.8355e+00 1.7135e+03 5.6005e-01
1249 4.996 -3.2585e+00 8.9739e-01 1.7135e+03 3.7785e-01
1250 5.000 -3.1534e+00 1.8295e-02 1.7135e+03 -4.5137e-01
1251 5.004 -3.5228e+00 4.7177e-02 1.7135e+03 6.9378e-01

12-3
 Taiwan Data: Examine, Zero Traces, Re-examine

The first column is time sample, the second column is time value (seconds). The next
four columns are amplitude values for traces 141-144. The amplitude values of trace 143
are constant at 1713.5 while the amplitudes of the other traces vary at lower amplitudes.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 12.2: Shot gather (fldr) 930 from file Tshot5.su.

As we examine the shot gathers with iview, we find that every gather has these same
three bad tracf traces; that is, we see the output of the same three bad hydrophones. In the
next section, we will zero (“kill”) the amplitude values of traces with tracf values 61, 62,
and 143.
Note: It is rare that traces are actually removed from a file. By keeping every trace in
our Taiwan shot gathers, every fldr continues to have exactly 144 traces. This regularity
can be useful in (is required by) some processing programs.

Figure 12.3: Shot gather (fldr) 820 from file Tshot5.su.

12-4
 Taiwan Data: Examine, Zero Traces, Re-examine

12.4 Kill Traces

The following lines are from the sukill selfdoc:
SUKILL - zero out traces
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

sukill <stdin >stdout min= count=1

Required parameters
min= first trace to kill (one-based)

Optional parameters
count= 1 number of traces to kill
Program sukill does not use a key; it finds traces by counting them. For our first shot
gather, we have to use sukill twice:
min=61 count=2
min=143 count=1
For our second shot gather, we have to use sukill twice:
min=205 count=2
min=287 count=1
We need 2 x 176 uses of sukill. Below is script killer.sh that generates text for sukill.
1 #! /bin/sh
2 # File: killer.sh
3 # Generate sukill values for Taiwan shots
4
5 # Set messages on
6 ##set -x
7
8 outtext=killer.txt
9 rm -f $outtext # remove earlier trials
10
11 increment=144 # number of traces per shot gather
12 firsts=800 # first fldr value
13 lasts=975 # last fldr value
14
15 bad2=61 # bad traces count 61,62
16 bad1=143 # bad trace count 143
17
18 #------------------------------------------------
19
20 k=0 # fldr counter
21 j=`expr $lasts - $firsts + 1`
22 i=1 # loop counter
23 while [ $i -le $j ]
24 do
25
26 two=`expr $k + $bad2`
27 one=`expr $k + $bad1`
28 echo "sukill min=$two count=2 | sukill min=$one count=1 |" >> $outtext
29
30 k=`expr $k + $increment` # fldr counter
31 i=`expr $i + 1` # loop counter
32
33 done
34
35 #------------------------------------------------
36
37 # Exit politely from shell
38 exit
39

12-5
 Taiwan Data: Examine, Zero Traces, Re-examine

Below is script killer2.sh that zeros the problem traces. Lines 11-186 were generated by
killer.sh (except that on line 11 we added “< $indata” and on line 186 we replaced
the pipe (|) with “> $outdata”.
1 #! /bin/sh
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

2 # File: killer2.sh
3 # kill regularly appearing bad traces
4
5 # Set messages on
6 set -x
7
8 indata=Tdata/Tshot5.su
9 outdata=Tdata/Tshot5k.su
10
11 sukill < $indata min=61 count=2 | sukill min=143 count=1 |
12 sukill min=205 count=2 | sukill min=287 count=1 |
13 sukill min=349 count=2 | sukill min=431 count=1 |
14 sukill min=493 count=2 | sukill min=575 count=1 |
15 sukill min=637 count=2 | sukill min=719 count=1 |
16 sukill min=781 count=2 | sukill min=863 count=1 |
17 sukill min=925 count=2 | sukill min=1007 count=1 |
18 sukill min=1069 count=2 | sukill min=1151 count=1 |
19 sukill min=1213 count=2 | sukill min=1295 count=1 |
20 sukill min=1357 count=2 | sukill min=1439 count=1 |
21 sukill min=1501 count=2 | sukill min=1583 count=1 |
22 sukill min=1645 count=2 | sukill min=1727 count=1 |
23 sukill min=1789 count=2 | sukill min=1871 count=1 |
24 sukill min=1933 count=2 | sukill min=2015 count=1 |
25 sukill min=2077 count=2 | sukill min=2159 count=1 |
26 sukill min=2221 count=2 | sukill min=2303 count=1 |
27 sukill min=2365 count=2 | sukill min=2447 count=1 |
28 sukill min=2509 count=2 | sukill min=2591 count=1 |
29 sukill min=2653 count=2 | sukill min=2735 count=1 |
30 sukill min=2797 count=2 | sukill min=2879 count=1 |
31 sukill min=2941 count=2 | sukill min=3023 count=1 |
32 sukill min=3085 count=2 | sukill min=3167 count=1 |
33 sukill min=3229 count=2 | sukill min=3311 count=1 |
34 sukill min=3373 count=2 | sukill min=3455 count=1 |
35 sukill min=3517 count=2 | sukill min=3599 count=1 |
36 sukill min=3661 count=2 | sukill min=3743 count=1 |
37 sukill min=3805 count=2 | sukill min=3887 count=1 |
38 sukill min=3949 count=2 | sukill min=4031 count=1 |
39 sukill min=4093 count=2 | sukill min=4175 count=1 |
40 sukill min=4237 count=2 | sukill min=4319 count=1 |
41 sukill min=4381 count=2 | sukill min=4463 count=1 |
42 sukill min=4525 count=2 | sukill min=4607 count=1 |
43 sukill min=4669 count=2 | sukill min=4751 count=1 |
44 sukill min=4813 count=2 | sukill min=4895 count=1 |
45 sukill min=4957 count=2 | sukill min=5039 count=1 |
46 sukill min=5101 count=2 | sukill min=5183 count=1 |
47 sukill min=5245 count=2 | sukill min=5327 count=1 |
48 sukill min=5389 count=2 | sukill min=5471 count=1 |
49 sukill min=5533 count=2 | sukill min=5615 count=1 |
50 sukill min=5677 count=2 | sukill min=5759 count=1 |
51 sukill min=5821 count=2 | sukill min=5903 count=1 |
52 sukill min=5965 count=2 | sukill min=6047 count=1 |
53 sukill min=6109 count=2 | sukill min=6191 count=1 |
54 sukill min=6253 count=2 | sukill min=6335 count=1 |
55 sukill min=6397 count=2 | sukill min=6479 count=1 |
56 sukill min=6541 count=2 | sukill min=6623 count=1 |
57 sukill min=6685 count=2 | sukill min=6767 count=1 |
58 sukill min=6829 count=2 | sukill min=6911 count=1 |
59 sukill min=6973 count=2 | sukill min=7055 count=1 |

12-6
 Taiwan Data: Examine, Zero Traces, Re-examine

60 sukill min=7117 count=2 | sukill min=7199 count=1 |

61 sukill min=7261 count=2 | sukill min=7343 count=1 |
62 sukill min=7405 count=2 | sukill min=7487 count=1 |
63 sukill min=7549 count=2 | sukill min=7631 count=1 |
64 sukill min=7693 count=2 | sukill min=7775 count=1 |
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

65 sukill min=7837 count=2 | sukill min=7919 count=1 |

66 sukill min=7981 count=2 | sukill min=8063 count=1 |
67 sukill min=8125 count=2 | sukill min=8207 count=1 |
68 sukill min=8269 count=2 | sukill min=8351 count=1 |
69 sukill min=8413 count=2 | sukill min=8495 count=1 |
70 sukill min=8557 count=2 | sukill min=8639 count=1 |
71 sukill min=8701 count=2 | sukill min=8783 count=1 |
72 sukill min=8845 count=2 | sukill min=8927 count=1 |
73 sukill min=8989 count=2 | sukill min=9071 count=1 |
74 sukill min=9133 count=2 | sukill min=9215 count=1 |
75 sukill min=9277 count=2 | sukill min=9359 count=1 |
76 sukill min=9421 count=2 | sukill min=9503 count=1 |
77 sukill min=9565 count=2 | sukill min=9647 count=1 |
78 sukill min=9709 count=2 | sukill min=9791 count=1 |
79 sukill min=9853 count=2 | sukill min=9935 count=1 |
80 sukill min=9997 count=2 | sukill min=10079 count=1 |
81 sukill min=10141 count=2 | sukill min=10223 count=1 |
82 sukill min=10285 count=2 | sukill min=10367 count=1 |
83 sukill min=10429 count=2 | sukill min=10511 count=1 |
84 sukill min=10573 count=2 | sukill min=10655 count=1 |
85 sukill min=10717 count=2 | sukill min=10799 count=1 |
86 sukill min=10861 count=2 | sukill min=10943 count=1 |
87 sukill min=11005 count=2 | sukill min=11087 count=1 |
88 sukill min=11149 count=2 | sukill min=11231 count=1 |
89 sukill min=11293 count=2 | sukill min=11375 count=1 |
90 sukill min=11437 count=2 | sukill min=11519 count=1 |
91 sukill min=11581 count=2 | sukill min=11663 count=1 |
92 sukill min=11725 count=2 | sukill min=11807 count=1 |
93 sukill min=11869 count=2 | sukill min=11951 count=1 |
94 sukill min=12013 count=2 | sukill min=12095 count=1 |
95 sukill min=12157 count=2 | sukill min=12239 count=1 |
96 sukill min=12301 count=2 | sukill min=12383 count=1 |
97 sukill min=12445 count=2 | sukill min=12527 count=1 |
98 sukill min=12589 count=2 | sukill min=12671 count=1 |
99 sukill min=12733 count=2 | sukill min=12815 count=1 |
100 sukill min=12877 count=2 | sukill min=12959 count=1 |
101 sukill min=13021 count=2 | sukill min=13103 count=1 |
102 sukill min=13165 count=2 | sukill min=13247 count=1 |
103 sukill min=13309 count=2 | sukill min=13391 count=1 |
104 sukill min=13453 count=2 | sukill min=13535 count=1 |
105 sukill min=13597 count=2 | sukill min=13679 count=1 |
106 sukill min=13741 count=2 | sukill min=13823 count=1 |
107 sukill min=13885 count=2 | sukill min=13967 count=1 |
108 sukill min=14029 count=2 | sukill min=14111 count=1 |
109 sukill min=14173 count=2 | sukill min=14255 count=1 |
110 sukill min=14317 count=2 | sukill min=14399 count=1 |
111 sukill min=14461 count=2 | sukill min=14543 count=1 |
112 sukill min=14605 count=2 | sukill min=14687 count=1 |
113 sukill min=14749 count=2 | sukill min=14831 count=1 |
114 sukill min=14893 count=2 | sukill min=14975 count=1 |
115 sukill min=15037 count=2 | sukill min=15119 count=1 |
116 sukill min=15181 count=2 | sukill min=15263 count=1 |
117 sukill min=15325 count=2 | sukill min=15407 count=1 |
118 sukill min=15469 count=2 | sukill min=15551 count=1 |
119 sukill min=15613 count=2 | sukill min=15695 count=1 |
120 sukill min=15757 count=2 | sukill min=15839 count=1 |
121 sukill min=15901 count=2 | sukill min=15983 count=1 |
122 sukill min=16045 count=2 | sukill min=16127 count=1 |

12-7
 Taiwan Data: Examine, Zero Traces, Re-examine

123 sukill min=16189 count=2 | sukill min=16271 count=1 |

124 sukill min=16333 count=2 | sukill min=16415 count=1 |
125 sukill min=16477 count=2 | sukill min=16559 count=1 |
126 sukill min=16621 count=2 | sukill min=16703 count=1 |
127 sukill min=16765 count=2 | sukill min=16847 count=1 |
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

128 sukill min=16909 count=2 | sukill min=16991 count=1 |

129 sukill min=17053 count=2 | sukill min=17135 count=1 |
130 sukill min=17197 count=2 | sukill min=17279 count=1 |
131 sukill min=17341 count=2 | sukill min=17423 count=1 |
132 sukill min=17485 count=2 | sukill min=17567 count=1 |
133 sukill min=17629 count=2 | sukill min=17711 count=1 |
134 sukill min=17773 count=2 | sukill min=17855 count=1 |
135 sukill min=17917 count=2 | sukill min=17999 count=1 |
136 sukill min=18061 count=2 | sukill min=18143 count=1 |
137 sukill min=18205 count=2 | sukill min=18287 count=1 |
138 sukill min=18349 count=2 | sukill min=18431 count=1 |
139 sukill min=18493 count=2 | sukill min=18575 count=1 |
140 sukill min=18637 count=2 | sukill min=18719 count=1 |
141 sukill min=18781 count=2 | sukill min=18863 count=1 |
142 sukill min=18925 count=2 | sukill min=19007 count=1 |
143 sukill min=19069 count=2 | sukill min=19151 count=1 |
144 sukill min=19213 count=2 | sukill min=19295 count=1 |
145 sukill min=19357 count=2 | sukill min=19439 count=1 |
146 sukill min=19501 count=2 | sukill min=19583 count=1 |
147 sukill min=19645 count=2 | sukill min=19727 count=1 |
148 sukill min=19789 count=2 | sukill min=19871 count=1 |
149 sukill min=19933 count=2 | sukill min=20015 count=1 |
150 sukill min=20077 count=2 | sukill min=20159 count=1 |
151 sukill min=20221 count=2 | sukill min=20303 count=1 |
152 sukill min=20365 count=2 | sukill min=20447 count=1 |
153 sukill min=20509 count=2 | sukill min=20591 count=1 |
154 sukill min=20653 count=2 | sukill min=20735 count=1 |
155 sukill min=20797 count=2 | sukill min=20879 count=1 |
156 sukill min=20941 count=2 | sukill min=21023 count=1 |
157 sukill min=21085 count=2 | sukill min=21167 count=1 |
158 sukill min=21229 count=2 | sukill min=21311 count=1 |
159 sukill min=21373 count=2 | sukill min=21455 count=1 |
160 sukill min=21517 count=2 | sukill min=21599 count=1 |
161 sukill min=21661 count=2 | sukill min=21743 count=1 |
162 sukill min=21805 count=2 | sukill min=21887 count=1 |
163 sukill min=21949 count=2 | sukill min=22031 count=1 |
164 sukill min=22093 count=2 | sukill min=22175 count=1 |
165 sukill min=22237 count=2 | sukill min=22319 count=1 |
166 sukill min=22381 count=2 | sukill min=22463 count=1 |
167 sukill min=22525 count=2 | sukill min=22607 count=1 |
168 sukill min=22669 count=2 | sukill min=22751 count=1 |
169 sukill min=22813 count=2 | sukill min=22895 count=1 |
170 sukill min=22957 count=2 | sukill min=23039 count=1 |
171 sukill min=23101 count=2 | sukill min=23183 count=1 |
172 sukill min=23245 count=2 | sukill min=23327 count=1 |
173 sukill min=23389 count=2 | sukill min=23471 count=1 |
174 sukill min=23533 count=2 | sukill min=23615 count=1 |
175 sukill min=23677 count=2 | sukill min=23759 count=1 |
176 sukill min=23821 count=2 | sukill min=23903 count=1 |
177 sukill min=23965 count=2 | sukill min=24047 count=1 |
178 sukill min=24109 count=2 | sukill min=24191 count=1 |
179 sukill min=24253 count=2 | sukill min=24335 count=1 |
180 sukill min=24397 count=2 | sukill min=24479 count=1 |
181 sukill min=24541 count=2 | sukill min=24623 count=1 |
182 sukill min=24685 count=2 | sukill min=24767 count=1 |
183 sukill min=24829 count=2 | sukill min=24911 count=1 |
184 sukill min=24973 count=2 | sukill min=25055 count=1 |
185 sukill min=25117 count=2 | sukill min=25199 count=1 |

12-8
 Taiwan Data: Examine, Zero Traces, Re-examine

186 sukill min=25261 count=2 | sukill min=25343 count=1 > $outdata

187
188 # Exit politely from shell
189 exit
190
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 12.4 shows shot gathers 930 and 820 after zeroing tracr traces 61, 62, and 143.
The ellipses on shot gather 820 show where 1-D frequency analysis is done in the next
section.

Figure 12.4: Shot (fldr) gathers 930 (top) and 820 (bottom) from file Tshot5k.su.
Note: In the Taiwan shot-order data, these bad traces have regular position. In CMP-
order data, these bad traces have no regular position. If we had not examined the gathers
before sorting them to CMP order, we would not have recognized the regularity of these
traces (that they were caused by particular bad hydrophones) and we would have to use a
different way to identify these traces. One other way to identify these traces is by their
regular high amplitudes.

12-9
 Taiwan Data: Examine, Zero Traces, Re-examine

12.5 One-dimensional Frequency Analysis

In Section 10.4.1, we used script fxdisp to see the frequencies of a gather. The
frequency display of fxdisp is the transform of each trace. Here, we present script tf.sh
that plots frequencies of a single trace or part of a single trace. Our examples use a half-
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

second of data from traces windowed from shot gather 820, shown by the ellipses in the
bottom shot gather of Figure 12.4.
In Figure 12.5, the left side example is from a low frequency noise area: tracf=55, 4-
4.5 seconds.
suwind < Tdata/Tshot5k.su key=fldr min=820 max=820 | suwind key=tracf min=55
max=55 tmin=4 tmax=4.5 > T820f55t4.su
We created the left side plots with the following command:
tf.sh T820f55t4.su
The right side example in Figure 12.5 is from a good signal area: tracf=93, 2-2.5
seconds.
suwind < Tdata/Tshot5k.su key=fldr min=820 max=820 | suwind key=tracf min=93
max=93 tmin=2 tmax=2.5 > T820f93t2.su
We created the right side plots with the following command:
tf.sh T820f93t2.su

Figure 12.5: 1-D frequency analysis of wavelets from shot gather 820. Left: low
frequency noise wavelet. Right: good high frequency wavelet.

12-10
 Taiwan Data: Examine, Zero Traces, Re-examine

Script tf.sh makes three plots. The top plot is the input time series, the middle plot is
the frequency transform of the time series, and the bottom plot is the phase spectrum of
the time series. Below is script tf.sh.
1 #! /bin/sh
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

2 # File: tf.sh
3 # Time to frequency transform of a trace
4 # Output = 3 plots: time, freq amplitude, phase spectrum
5 # Use: tf.sh [input.su]
6 # Example: tf.sh wave1.su
7
8 # Set messages on
9 set -x
10
11 suxgraph < $1 -geometry 400x200+10+10 -bg white \
12 title="Input: $1" \
13 label1="Time (s)" label2="Amplitude" \
14 style=normal linecolor=0 &
15
16 ##suxwigb < $1 wt=1 va=2 style=normal labelcolor=black \
17 ## wbox=400 hbox=200 xbox=420 ybox=10 \
18 ## title="Input: $1" titlecolor=black \
19 ## label1="Time (s)" label2=" Amplitude" &
20
21 sufft < $1 | suamp mode=amp |
22 suxgraph -geometry 400x200+10+245 -bg white \
23 title="Amplitude Spectrum" \
24 label1="Frequency (Hz)" label2="Amplitude" \
25 style=normal linecolor=0 &
26
27 sufft < $1 | suamp mode=phase |
28 suxgraph -geometry 400x200+10+480 -bg white \
29 title="Phase Spectrum" \
30 label1="Frequency (Hz)" label2="Phase (rad)" \
31 style=normal linecolor=0 &
32
33 # "press return key to ..."
34 pause remove the plots
35
36 zap xgraph
37 ##zap xwigb
38
39 # Exit politely from shell
40 exit
41
• Line 6: As the example shows, the input file must be supplied when the script is
used. The input file is variable $1.
• Lines 11-14 make the time series plot with suxgraph.
• Line 21: sufft transforms the data and suamp mode=amp outputs the amplitude
spectrum of the information from sufft.
• Lines 22-25 plot the amplitude spectrum with suxgraph.
• Line 27: sufft transforms the data and suamp mode=phase outputs the phase
spectrum of the information from sufft.
• Lines 28-31 plot the phase spectrum with suxgraph.
• Lines 16-19 make an alternative time series plot using suxwigb. These lines can
be uncommented for use.

12-11
 Taiwan Data: Examine, Zero Traces, Re-examine

Note: If you middle click the mouse in an xwigb window, mouse location
information is printed in the upper left of the window. An xgraph window does not have
this feature.
• After the plots are made, Line 34 causes the following line to be printed to the
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

screen: press return key to remove the plots

Script tf.sh is a display script; no data are output.

12-12
 Taiwan: Gain-Filter, Filter-Gain

13. Taiwan: Gain-Filter, Filter-Gain

13.1 Introduction
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

We now have 5 seconds of Taiwan data and we zeroed the bad hydrophones. In this
chapter we will
1. apply spherical divergence correction, then band-pass filter the data.
We will also
2. apply a band-pass filter, then apply spherical divergence correction.
We will compare gathers from these two paths and consider whether one path is better.

13.2 Gain-Filter

13.2.1 Gain – Spherical Divergence Correction

To determine which exponential gain to apply to the line of CMPs, we will apply gain
to two shot gathers: 820 and 930.. We window them from the 2-D line:
suwind < Tdata/Tshot5k.su key=fldr min=820 max=820 > T5k820.su
suwind < Tdata/Tshot5k.su key=fldr min=930 max=930 > T5k930.su
We will determine a tpow value for sugain by using script igain (Section 10.5),
option T. Below is the user area of igain.sh changed for the Taiwan data.
9 # User-supplied values
10
11 indata=T5k820.su # Input file
12 myperc=90 # perc value
The script creates three displays:
• a wiggle plot of the file,
• a decibel (dB) ximage plot of the user-selected trace(s), and
• an amplitude graph of the user-selected trace(s).
Remember:
• The script applies gain to the entire file, but only the user-selected trace or traces
are used to create the second and third displays.
• The user is asked to supply the name of a key (offset, tracr, etc.) and
corresponding key minimum and key maximum values for the second and third
displays.
We selected key tracf and supplied tracf values 92 93 for both shot gathers, based on our
displays in Chapter 12.
Table 13.1: dB ranges of Figures 13.1 and 13.2
fldr tracf No gain tpow = 1.8
maximum dB 0 0
820 92 93
minimum dB -79.4 -110.6
maximum dB 0 0
930 92 93
minimum dB -82.53 -123.9

13-1
 Taiwan: Gain-Filter, Filter-Gain

After several trials, we decide to use a spherical divergence correction of 1.8 (Figures
13.1 and 13.2).
We apply spherical divergence correction with the command below.
sugain < Tdata/Tshot5k.su tpow=1.8 > Tdata/Tshot5kg.su
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 13.1: Top: Shot 820 without gain. Bottom: Shot 820 with gain, tpow=1.8.

13-2
 Taiwan: Gain-Filter, Filter-Gain
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 13.2: Top: Shot 930 without gain. Bottom: Shot 930 with gain, tpow=1.8.

13.2.2 Band-pass Filter

To select frequencies for band-pass filtering, we window two shot gathers from the
gained 2-D line and examine them with script fxdisp.

13-3
 Taiwan: Gain-Filter, Filter-Gain

suwind < Tdata/Tshot5kg.su key=fldr min=820 max=820 > T5kg820.su

suwind < Tdata/Tshot5kg.su key=fldr min=930 max=930 > T5kg930.su
The fxdisp dialog asks for three values:
• the input file name,
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

• a perc value for the wiggle plot of the input file, and
• a key (trace header) to label the x-axis of the wiggle plot.
The script creates two displays: a wiggle plot of the input traces and a corresponding
frequency display. The script transforms each input seismic trace to a frequency trace.
The script repeats these three questions and two displays until you quit the script. No
data are saved after processing.
If you make a mistake while typing, use the Delete key, not the Back Space key.

Figure 13.3: Plots of f-x and t-x after gain. Top: Shot 820. Bottom: Shot 930.
The top of Figure 13.3 shows that the low frequency noise overwhelms the
reflections.

13-4
 Taiwan: Gain-Filter, Filter-Gain

Note: The x-axis annotation of the frequency plot (an xwigb window) shows tracr
values, the first default key. See Section 7.7.
Based on Figure 13.3, we will use the following command to filter the gained data:
sufilter < Tdata/Tshot5kg.su f=18,23,55,60 amps=0,1,1,0 > Tdata/Tshot5kgf.su
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

13.2.3 Gain-Filter Result

Figure 13.4 shows our two test shot gathers after spherical divergence correction
followed by our band-pass filter.

Figure 13.4: After gain, then band-pass filter. Top: Shot 820. Bottom: Shot 930.
Let’s continue our experiments by reversing the process order.

13-5
 Taiwan: Gain-Filter, Filter-Gain

13.3 Filter-Gain

13.3.1 Band-pass Filter

To select frequencies for band-pass filtering, we examine the two shot gathers we
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

windowed from the 2-D line at the beginning of Section 13.2: T5k820.su and T5k930.su.
We examine these files with script fxdisp.

Figure 13.5: Plots of f-x and t-x. Top: Shot 820. Bottom: Shot 930.
Compare Figure 13.5, no gain, to Figure 13.3, fxdisp after gain. Without gain, the low
frequency noise on the near-offset traces of gather 820 does not overwhelm the
reflections.
After discussion, we decide to use the same sufilter values we used in Section 13.3.
sufilter < Tdata/Tshot5k.su f=18,23,55,60 amps=0,1,1,0 > Tdata/Tshot5kf.su

13.3.2 Gain – Spherical Divergence Correction

We will apply exponential gain to two shot gathers windowed from the filtered data:

13-6
 Taiwan: Gain-Filter, Filter-Gain

suwind < Tdata/Tshot5kf.su key=fldr min=820 max=820 > T5kf820.su

suwind < Tdata/Tshot5kf.su key=fldr min=930 max=930 > T5kf930.su
We will determine a tpow value for sugain by using script igain, option T. Again, we
will use key tracf and tracf values 92 93.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 13.6: Top: Shot 820 without gain. Bottom: Shot 820 with gain, tpow=2.0.

13-7
 Taiwan: Gain-Filter, Filter-Gain

This time, we think a tpow value of 2.0 nicely balances the data (Figures 13.6, 13.7).
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 13.7: Top: Shot 930 without gain. Bottom: Shot 930 with gain, tpow=2.0.
Table 13.2 shows the amplitude range increase of spherical divergence correction.
We apply spherical divergence correction with the command below.

13-8
 Taiwan: Gain-Filter, Filter-Gain

sugain < Tdata/Tshot5kf.su tpow=2.0 > Tdata/Tshot5kfg.su

Table 13.2: dB ranges of Figures 13.5 and 13.6
fldr tracf No gain tpow = 2.0
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

maximum dB 0 0
820 92 93
minimum dB -94 -146.9
maximum dB 0 0
930 92 93
minimum dB -86.59 -151.5

13.3.3 Filter-Gain Result

Figure 13.8 shows our two test shot gathers after band-pass filtering followed by
spherical divergence correction.

Figure 13.8: After band-pass filter, then gain. Top: Shot 820. Bottom: Shot 930.

13-9
 Taiwan: Gain-Filter, Filter-Gain

13.4 Discussion
When we compare Figures 13.4 and 13.8, we see minor differences, but not major
differences. Recall that we used different gain values in the two tests. While we used the
same frequency pass-band for both tests, we did consider passing slightly more lower
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

frequencies in the second test (Section 13.3.1).

Further analysis can be done using sudiff to get a difference file. For example:
sudiff T5kgf820.su T5kfg820.su > Tdiff820.su
sudiff T5kgf930.su T5kfg930.su > Tdiff930.su
To see the documentation for sudiff, enter “sudiff” or “sudoc sudiff”.
Also, you can examine trace amplitudes by using suascii or sudumptrace. Program
sudumptrace is described in Appendix C.

13-10
 f-k Filter and Deconvolution

14. f-k Filter and Deconvolution

14.1 Introduction
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Real data have noise of various kinds. Some signal enhancement programs are used
prestack (before stacking) and some are used post-stack (after stacking CMPs). In the
previous chapter, we used gain and band-pass filtering on prestack data. We only used
one option out of many in sugain, but you are free to test other sugain options. In this
chapter, we present two popular signal enhancement tools. These tools could prove useful
in processing the Taiwan data set; however, we were unable to improve that data set
using them. (Perhaps you can find a way!). To demonstrate their capability, we use them
on Oz files (shot gathers) that we get from the Colorado School of Mines Center for
Wave Phenomena web site (Section 3.3).
The first tool is a frequency-wavenumber (f-k) filter, a two-dimensional filter. We see
how to use an f-k filter through the sudipfilt program. For a brief description of the f-k
domain, see Sheriff (2002): f-k domain.
The second tool is deconvolution. We implement a deconvolution (decon) script that
uses suacor, an autocorrelation program, and supef, a prediction-error filter.

14.2 f-k Filter -- Absolute Units

In Section 10.4.1, we used the one-dimensional transform display script fxdisp. Script
fxdisp transforms each trace from a time series to a frequency series. Here, we use
interactive script ifk to apply a two-dimensional frequency filter. The f-k filter slopes can
be specified in relative units of time over trace spacing (for example, seconds/meter) or in
absolute units of time samples per trace. In this section, we specify the f-k filter slopes in
absolute units: we measure the vertical axis as “number of samples” and we measure the
horizontal axis as “number of traces.”

14.2.1 Use the f-k Filter Script

Script ifk interactively lets us first view the f-k spectrum of a gather, then specify the
slopes of the pass/reject filter. In this example, the input file is Oz gather 8. We get Oz
gather 8, a split-spread gather, from the Colorado School of Mines web site. Below is the
surange output of oz08.su.
surange < oz08.su
96 traces:
tracl=(1,96) tracr=(1,96) fldr=10008 tracf=(1,96) cdp=(8,103)
cdpt=1 trid=1 nvs=1 nhs=1 duse=1
scalel=1 scalco=1 counit=1 delrt=4 muts=4
ns=1350 dt=4000
However, we want to use only one side of the split-spread gather. Also, oz08.su has 5.4
seconds; we will window that to the first 3 seconds.
suwind < oz08.su key=tracl min=26 max=96 tmax=3 > oz08w.su
Below is the surange output of oz08w.su.
surange < oz08w.su
71 traces:
tracl=(26,96) tracr=(26,96) fldr=10008 tracf=(26,96) cdp=(33,103)
cdpt=1 trid=1 nvs=1 nhs=1 duse=1

14-1
 f-k Filter and Deconvolution

scalel=1 scalco=1 counit=1 delrt=4 muts=4

ns=750 dt=4000
Below is the screen dialog with our input highlighted. Line numbers are added for
discussion. We input values on lines 14, 21, and 27. Note that our slope values, supplied
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

on dialog line 21, are 4,5,6,7 (absolute units).

1 ------------------------------------------------------
2 f-k Filter Test
3 ---------------
4 From your slope values, two filters are created --
5 a pass filter and a reject filter.
6 When you exit, the following files are output:
7 Slopes are in ==> fk.txt
8 Passed data are in ==> fkpass.su
9 Rejected data are in ==> fkrejj.su
10 ------------------------------------------------------
11
12 Supply gain power value for t^(power)
13 For no gain, supply 0
14 2
15
16 Supply filter slopes.
17 Input: s1,s2,s3,s4 where s1 < s2 < s3 < s4
18 Example: 3.0,3.5,4.0,4.5
19 or: -4.5,-4.0,-3.5,-3.0
20 Use commas. Do not use spaces.
21 4,5,6,7
22

Figure 14.1: Input gather oz08w.su. Left: Seismic after gain (tpow=2.0). Right: f-k plot
of seismic after gain.
23 Slopes: 4,5,6,7
24
25 Enter 1 for more f-k filter testing

14-2
 f-k Filter and Deconvolution

26 Enter 2 to EXIT
27 2
28
29 Slopes are in ==> fk.txt
30 Passed data are in ==> fkpass.su
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

31 Rejected data are in ==> fkrejj.su

32
33 press return key to exit
The script lets us apply a t^(power) gain to the input seismic data (dialog lines 12-14).
After that question is answered, the first displays are made. One shows the input seismic
data, the other shows an f-k plot of the seismic data (Figure 14.1).
Using the four user-supplied slope values (dialog lines 16-20), script ifk makes two
outputs using sudipfilt:
• a pass filter
slopes=s1,s2,s3,s4 amps=0,1,1,0
• a reject filter
slopes=s1,s2,s3,s4 amps=1,0,0,1
Figure 14.2 shows the four output plots: a t-x plot and an f-k plot after application of
the pass filter and a t-x plot and an f-k plot after application of the reject filter. The test
can be repeated. When the user exits the script, the seismic results of the last test are
output: the pass seismic data (fkpass.su) and the reject seismic data (fkrejj.su). A log of
the processing steps is output as file fk.txt (shown below).
f-k Filter Test
---------------
sugain < oz08w.su tpow=2
==> First test
Slopes: 4,5,6,7
sudipfilt < oz08w.su dx=1 dt=1 \
slopes= amps=
Passed data are in ==> fkpass.su
Rejected data are in ==> fkrejj.su
In output file fk.txt, no values are output for parameter amps because it is left to the
user whether to use slopes as a pass filter (amps=0,1,1,0) or as a reject filter
(amps=1,0,0,1).
We rename the output files to prevent our files from becoming overwritten by a later
run: Note: Use a semicolon to separate multiple commands on the same line.
mv fkpass.su fkpass08abs.su; mv fkrejj.su fkrejj08abs.su

14.2.2 Script ifk.scr

We usually start interactive scripts by running a .scr script. Here is ifk.scr:
1 #! /bin/sh
2 # File: ifk.scr
3 # Run this script to start script ifk.sh
4
5 xterm -geom 80x17+10+640 -e ifk.sh
6
Line 1 invokes the shell. Line 5 opens a dialog window 80 characters wide by 17
characters high. Line 5 also places the dialog window 10 pixels from the left side of the
viewing area and 640 pixels from the top of the viewing area. Line 5 also starts ifk.sh, the
processing script, within that window.

14-3
 f-k Filter and Deconvolution
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 14.2: Output of slope values 4,5,6,7. Top: Data passed by filter. Bottom: Data
rejected by filter.

14-4
 f-k Filter and Deconvolution

14.2.3 Script ifk.sh

We start interactive scripts by running a .scr script. Here is ifk.sh.
1 #! /bin/sh
2 # File: ifk.sh
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

3 # Run script ifk.scr to start this script

4 # Interactive f-k processing
5
6 # Set messages on
7 ##set -x
8
9 #================================================
10 # USER AREA -- SUPPLY VALUES
11 #------------------------------------------------
12
13 # Input seismic file
14 indata=oz08w.su
15
16 # Display choices
17 myperc=99 # perc value for seismic plots
18 plottype=0 # 0 = wiggle plot, 1 = image plot
19
20 # Processing variables [ Instructions below ]
21 dx=1 # trace spacing
22 dt=1 # time sample interval
23
24 #================================================
25
26 # Instructions:
27 # ------------
28 # sudipfilt is a classic pie slice filter. Each slope radiates
29 # from the f-k origin. Supply four slope values. A "slope" is
30 #
31 # change in t
32 # -----------
33 # change in x
34 #
35 # You can use relative units: (seconds/meter)
36 # for example: dt=.004 dx=25
37 # or absolute units: (time samples / trace)
38 # for example: dt=1 dx=1
39 #
40 #------------------------------------------------
41
42 echo " ------------------------------------------------------"
43 echo " f-k Filter Test"
44 echo " ---------------"
45 echo " From your slope values, two filters are created --"
46 echo " a pass filter and a reject filter."
47 echo " When you exit, the following files are output:"
48 echo " Slopes are in ==> fk.txt"
49 echo " Passed data are in ==> fkpass.su"
50 echo " Rejected data are in ==> fkrejj.su"
51 echo " ------------------------------------------------------"
52
53 # Remove temporary files
54 rm -f tmp*
55
56 #------------------------------------------------
57 # Describe temporary files
58 #------------------------------------------------
59
60 # tmp0 = Binary. A copy of the input seismic file

14-5
 f-k Filter and Deconvolution

61 # tmp1 = Binary. Data that is processed, either tmp0 or tmp2

62 # tmp1 is a copy of tmp0 if re-processing not wanted
63 # tmp1 is a copy of tmp2 if re-processing is wanted
64 # tmp2 = Binary. Output of "a=0,1,1,0" (pass) dip filtering
65 # tmp3 = Binary. Output of "a=1,0,0,1" (reject) dip filtering
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

66 # tmp4 = ASCII record of dialog for output text file

67 # tmp5 = ASCII file to reduce "zap" screen messages
68
69 #------------------------------------------------
70 # Supply gain value to t^(power)
71 #------------------------------------------------
72
73 echo " "
74 echo "Supply gain power value for t^(power)"
75 echo "For no gain, supply 0"
76 > /dev/tty
77 read tpow
78
79 # Copy input data to temporary file; apply gain if requested
80 sugain < $indata tpow=$tpow > tmp0
81
82 # Log input file & gain value
83 echo " f-k Filter Test" > tmp4
84 echo " ---------------" >> tmp4
85 echo "sugain < $indata tpow=$tpow" >> tmp4
86
87 #------------------------------------------------
88 # Plot original gather and spectrum
89 #------------------------------------------------
90
91 if [ $plottype -eq 0 ] ; then
92 suxwigb < tmp0 xbox=10 ybox=10 wbox=300 hbox=500 \
93 label1=" Time (s)" label2="Offset" \
94 title="Original data; gain = $tpow" key=offset \
95 perc=$myperc verbose=0 &
96 else
97 suximage < tmp0 xbox=10 ybox=10 wbox=300 hbox=500 \
98 label1=" Time (s)" \
99 title="Original data; gain = $tpow" \
100 perc=$myperc verbose=0 &
101 fi
102
103 suspecfk < tmp0 dx=$dx dt=$dt |
104 suximage xbox=320 ybox=10 wbox=300 hbox=500 \
105 label1=" Frequency (Hz)" label2="Wavenumber (k)"\
106 title="f-k spectrum, no filter" \
107 cmap=hsv2 legend=1 units=Amplitude verbose=0 \
108 grid1=dots grid2=dots perc=99 &
109
110 #------------------------------------------------
111 # f-k filter test
112 #------------------------------------------------
113
114 new=true # true = first test
115 ok=false # false = continue looping
116
117 while [ $ok = false ]
118 do
119
120 rm -f tmp1 # remove earlier copy of file to be filtered
121
122 if [ $new = true ] ; then
123 echo " ==> First test" >> tmp4

14-6
 f-k Filter and Deconvolution

124 cp tmp0 tmp1

125 else
126 echo " "
127 echo "Enter A or a to add an f-k filter"
128 echo "Enter S or s to start over"
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

129 > /dev/tty

130 read choice1
131
132 case $choice1 in
133 [sS])
134 cp tmp0 tmp1
135 echo " ==> Using original data"
136 echo " ==> Using original data" >> tmp4
137 ;;
138 [aA])
139
140 echo " "
141 echo "Enter P or p to use Passed data"
142 echo "Enter R or r to use Rejected data"
143 > /dev/tty
144 read choice3
145
146 case $choice3 in
147 [pP])
148 echo " "
149 cp tmp2 tmp1
150 echo " ==> Using passed data"
151 echo " ==> Using passed data" >> tmp4
152 ;;
153 [rR])
154 echo " "
155 cp tmp3 tmp1
156 echo " ==> Using rejected data"
157 echo " ==> Using rejected data" >> tmp4
158 ;;
159 esac
160
161 ;;
162 esac
163
164 # Remove earlier test of passed and rejected data
165 rm -f tmp2
166 rm -f tmp3
167
168 fi
169
170 #------------------------------------------------
171 # Get filter slope values
172 #------------------------------------------------
173
174 echo " "
175 echo "Supply filter slopes."
176 echo "Input: s1,s2,s3,s4 where s1 < s2 < s3 < s4"
177 echo " Example: 3.0,3.5,4.0,4.5"
178 echo " or: -4.5,-4.0,-3.5,-3.0"
179 echo "Use commas. Do not use spaces."
180 > /dev/tty
181 read slopes
182 echo " "
183 echo " Slopes: $slopes"
184 echo " Slopes: $slopes" >> tmp4
185
186 #------------------------------------------------

14-7
 f-k Filter and Deconvolution

187 # Apply filters, plot seismic data, plot f-k data

188 #------------------------------------------------
189
190 # Apply pass filter
191 sudipfilt < tmp1 dx=$dx dt=$dt slopes=$slopes amps=0,1,1,0 > tmp2
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

192
193 # Plot seismic passed data
194 if [ $plottype -eq 0 ] ; then
195 suxwigb < tmp2 xbox=10 ybox=10 wbox=300 hbox=500 \
196 label1=" Time (s)" label2="Offset"\
197 title="After f-k PASS filter" key=offset \
198 perc=$myperc verbose=0 &
199 else
200 suximage < tmp2 xbox=10 ybox=10 wbox=300 hbox=500 \
201 label1=" Time (s)" \
202 title="After f-k PASS filter" \
203 perc=$myperc verbose=0 &
204 fi
205
206 # Plot f-k passed data
207 suspecfk < tmp2 dx=$dx dt=$dt |
208 suximage xbox=320 ybox=10 wbox=300 hbox=500 \
209 label1=" Frequency (Hz)" label2="Wavenumber (k)" \
210 title="f-k Spectrum after PASS filter" \
211 cmap=hsv2 legend=1 units=Amplitude verbose=0 \
212 grid1=dots grid2=dots perc=99 &
213
214 # Apply reject filter
215 sudipfilt < tmp1 dx=$dx dt=$dt slopes=$slopes amps=1,0,0,1 > tmp3
216
217 # Plot seismic rejected data
218 if [ $plottype -eq 0 ] ; then
219 suxwigb < tmp3 xbox=630 ybox=10 wbox=300 hbox=500 \
220 label1=" Traveltime (s)" label2="Offset" \
221 title="After f-k REJECT filter" key=offset \
222 perc=$myperc verbose=0 &
223 else
224 suximage < tmp3 xbox=630 ybox=10 wbox=300 hbox=500 \
225 label1=" Time (s)" \
226 title="After f-k REJECT filter" \
227 perc=$myperc verbose=0 &
228 fi
229
230 # Plot f-k rejected data
231 suspecfk < tmp3 dx=$dx dt=$dt |
232 suximage xbox=940 ybox=10 wbox=300 hbox=500 \
233 label1=" Frequency (Hz)" label2="Wavenumber (k)" \
234 title="f-k Spectrum after REJECT filter" \
235 cmap=hsv2 legend=1 units=Amplitude verbose=0 \
236 grid1=dots grid2=dots perc=99 &
237
238 #------------------------------------------------
239 # More f-k or exit
240 #------------------------------------------------
241
242 echo " "
243 echo "Enter 1 for more f-k filter testing"
244 echo "Enter 2 to EXIT"
245 > /dev/tty
246 read choice2
247
248 case $choice2 in
249 1)

14-8
 f-k Filter and Deconvolution

250 ok=false
251 ;;
252 2)
253 cp tmp2 fkpass.su
254 cp tmp3 fkrejj.su
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

255 echo "sudipfilt < $indata dx=$dx dt=$dt \\" >> tmp4
256 echo " slopes= amps=" >> tmp4
257 echo " "
258 echo "Processing log is in ==> fk.txt"
259 echo " Passed data are in ==> fkpass.su"
260 echo " Passed data are in ==> fkpass.su" >> tmp4
261 echo "Rejected data are in ==> fkrejj.su"
262 echo "Rejected data are in ==> fkrejj.su" >> tmp4
263 cp tmp4 fk.txt
264 pause exit
265 zap xwigb > tmp5
266 zap ximage > tmp5
267 ok=true
268 ;;
269 esac
270
271 new=false # true = first test
272
273 done
274
275 #------------------------------------------------
276 # Exit
277 #------------------------------------------------
278
279 # Remove temporary files
280 rm -f tmp*
281
282 # Exit politely from shell
283 exit
284
The user-supplied values are on lines 13-22. Lines 21 and 22 are where trace spacing
and the time sample interval are set. The next lines, 26-40, discuss these parameters.
Lines 42-51 are written to the screen as soon as the script starts so the user will know
what outputs to expect.
Line 54 removes temporary files that might be left from a previously crashed run.
Lines 60-67 document the use of internal files.
Lines 73-77 ask the user if t^(power) gain is desired.
Line 80 copies the seismic data to an internal file while applying gain, or no gain.
Lines 83-85 write the input file name and the gain value to the temporary log files.
Lines 91-101 plot the input seismic data using suxwigb or suximage.
Line 103 transforms the input seismic data to f-k space. Lines 104-108 plot the 2-D
transform of the input seismic data.
Actual filter testing occurs between lines 114 and 273.
Line 114, parameter new, is a test for the if block of lines 122-168. If this is a first test
(the first time through the loop), the input data are copied from tmp0 to tmp1 (line 124).
If this is not the first test, the user is faced with choice1, lines 127-130. If the user
chooses to start fresh, the original data in tmp0 are copied to tmp1 (line 134). If the user
chooses to re-filter data already filtered, another choice (choice3) must be made: whether

14-9
 f-k Filter and Deconvolution

to reprocess the data from the “pass” filter or from the “reject” filter (lines 141-144).
After a previous test, the “pass” data are in tmp2 (line 191) and the “reject” data are in
tmp3 (line 215). The user’s choice (choice3) determines which file is copied to
(overwrites) tmp1 for the next test (line 149 or line 155). At the bottom of the test loop,
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

line 273, new=false insures that, after the first time through the loop, the user will always
have to see choice1.
If this is not the first time through the loop, lines 165-166 remove the previous “pass”
and “reject” data files.
Lines 175-181 explain the input format of the slopes and read the slopes from the
user.
Line 191 applies the “pass” filter (amps=0,1,1,0). Lines 194-204 plot the seismic
“pass” data. Line 207 transforms the “pass” seismic data to f-k space. Lines 208-212 plot
the 2-D transform of the “pass” seismic data.
Line 215 applies the “reject” filter (amps=1,0,0,1). Lines 218-228 plot the
seismic “reject” data. Line 231 transforms the “reject” seismic data to f-k space. Lines
232-236 plot the 2-D transform of the “reject” seismic data.
Lines 243-246 offer the user the choice (choice2) to re-filter or exit the script.
If the user chooses to re-filter (choice2=1), line 250 lets the script pass the user to the
top of the loop. If the user chooses to exit (choice2=2):
• Lines 253-254 create permanent disk files from the last processed “pass” and
“reject” test.
• Lines 255-256 put a template of the sudipfilt parameters into the temporary log
file.
• Lines 258-262 write information to the screen and the log file.
• Line 263 copies the temporary log file to a permanent disk file.
• Line 264 makes the script pause so the user can read the screen messages.
• Lines 265-266 close the plot windows and re-direct some of the accompanying
system messages to a temporary file.
• Line 267 sets a flag (ok=true) that ends cycling through the loop (see line 117).
After using ifk the first time, line 271 insures that the user is directed to the else
portion of if block lines 122-168.
Line 280 removes all temporary files.
Line 283 exits the script.

14.3 f-k Filter -- Relative Units

We said earlier that the f-k filter slopes can be specified in relative units of time over
trace spacing (for example, seconds/meter) or in absolute units of time samples per trace.
In this section, we specify the f-k filter slopes in relative units: we measure the vertical
axis as time in seconds and we measure the horizontal axis as distance in meters.
Again, we use time- and trace-windowed Oz gather 8. Below is the surange output of
oz08w.su.
surange < oz08w.su

14-10
 f-k Filter and Deconvolution

71 traces:
tracl=(26,96) tracr=(26,96) fldr=10008 tracf=(26,96) cdp=(33,103)
cdpt=1 trid=1 nvs=1 nhs=1 duse=1
scalel=1 scalco=1 counit=1 delrt=4 muts=4
ns=750 dt=4000
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

From surange, we know the sample interval is 0.004 seconds (dt=4000). From
Table 1-8 of Seismic Data Processing (Yilmaz, 1987) or Table 1-13 of Seismic Data
Analysis (Yilmaz, 2001), we know the trace spacing is 50 meters.
Table 14.1: Absolute and relative filter slope values
Absolute samples/trace 4 5 6 7
0.016/50 0.020/50 0.024/50 0.028/50
Relative seconds/meter
0.00032 0.00040 0.00048 0.00056
In script ifk.sh, we change lines 21-22:
20 # Processing variables [ Instructions below ]
21 dx=50 # trace spacing
22 dt=0.004 # time sample interval
Below is the screen dialog with our input highlighted. Line numbers are added for
discussion. We input values on lines 14, 21, and 27. We supply our slope values (relative
units) on dialog line 21.
1 ------------------------------------------------------
2 f-k Filter Test
3 ---------------
4 From your slope values, two filters are created --
5 a pass filter and a reject filter.
6 When you exit, the following files are output:
7 Slopes are in ==> fk.txt
8 Passed data are in ==> fkpass.su
9 Rejected data are in ==> fkrejj.su
10 ------------------------------------------------------
11
12 Supply gain power value for t^(power)
13 For no gain, supply 0
14 2
15
16 Supply filter slopes.
17 Input: s1,s2,s3,s4 where s1 < s2 < s3 < s4
18 Example: 3.0,3.5,4.0,4.5
19 or: -4.5,-4.0,-3.5,-3.0
20 Use commas. Do not use spaces.
21 .00032,.00040,.00048,.00056
22
23 Slopes: .00032,.00040,.00048,.00056
24
25 Enter 1 for more f-k filter testing
26 Enter 2 to EXIT
27 2
28
29 Processing log is in ==> fk.txt
30 Passed data are in ==> fkpass.su
31 Rejected data are in ==> fkrejj.su
32

14-11
 f-k Filter and Deconvolution

33 press return key to exit

Figure 14.3 shows the input time gather and its f-k equivalent.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 14.3: Input gather oz08w.su. Left: Seismic after gain (tpow=2.0). Right: f-k plot
of seismic after gain.
Figure 14.4 shows the four output plots: a t-x plot and an f-k plot after application of
the pass filter and a t-x plot and an f-k plot after application of the reject filter. (The test
can be repeated).
We rename the output files to prevent our files from becoming overwritten by a later
run: Note: Use a semicolon to separate multiple commands on the same line.
mv fkpass.su fkpass08rel.su; mv fkrejj.su fkrejj08rel.su
We are happy to see that Figure 14.2 and Figure 14.4 look the same. We compare
these data sets in the next section.

14-12
 f-k Filter and Deconvolution
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 14.4: Output of slope values 0.00032,0.00040,0.00048,0.00056. Top: Data

passed by filter. Bottom: Data rejected by filter.

14-13
 f-k Filter and Deconvolution

14.4 f-k Filter -- Discussion

First, we discuss the amplitudes of the two methods used above. Second, we discuss
the vertical and horizontal annotations on the f-k plots.
Starting with the reject files from above, we made a difference file:
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

sudiff fkrejj08abs.su fkrejj08rel.su > fkrejj08diff.su

Then, from all three data sets, we windowed times 1.000 to 1.040 from trace 50. The
amplitudes of the two methods are very similar, and the difference amplitudes are much
smaller than those of the absolute and relative files.
Table 14.2: Reject files, trace 50
Time Absolute Relative Difference
1.000 2.1742e+06 2.1742e+06 4.0000e+00
1.004 1.9120e+06 1.9120e+06 2.8750e+00
1.008 1.1568e+06 1.1568e+06 1.0000e+00
1.012 -8.8561e+05 -8.8561e+05 -5.0000e-01
1.016 -2.5479e+06 -2.5479e+06 -2.0000e+00
1.020 -2.0101e+06 -2.0101e+06 -3.1250e+00
1.024 -4.7880e+05 -4.7880e+05 -4.2812e+00
1.028 1.9655e+03 1.9688e+03 -3.2500e+00
1.032 -3.0930e+05 -3.0930e+05 -4.2500e+00
1.036 -4.5027e+05 -4.5026e+05 -3.3125e+00
1.040 -1.5578e+05 -1.5577e+05 -3.3125e+00
If we plot only the difference file, it is difficult to see that its amplitudes are much
smaller than those of the absolute and relative files.

Figure 14.5: File fkrejj08diff.su.

14-14
 f-k Filter and Deconvolution

Using the following commands, we concatenate the three files with the difference file
in the middle. The first command creates the file that eventually holds all three files.
cat fkrejj08abs.su > fkrejj08BIG.su
cat fkrejj08diff.su >> fkrejj08BIG.su
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

cat fkrejj08rel.su > fkrejj08BIG.su

Then we plot the result.
suxwigb < fkrejj08BIG.su perc=99 &

Figure 14.6: File fkrejj08BIG.su, the concatenation of the absolute method reject file
(left),the relative method reject file (right), and the difference of the two (middle).
In Figure 14.6, we see that the amplitudes of the difference file are much smaller than
the amplitudes of the other two files. Figure 14.6 shows vertical bands because the
computer screen cannot clearly show many traces squeezed into a narrow space.

14-15
 f-k Filter and Deconvolution

To understand the vertical and horizontal annotations on the f-k plots, remember the
definition of the Nyquist frequency. The Nyquist frequency, introduced in Section 10.4,
is the folding frequency. Mathematically, it is the inverse of twice the sample interval.
Units of temporal frequency (f) are the number of cycles per second. Units of spatial
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

frequency (wavenumber or k) are the number of cycles per unit distance. In Table 14.3,
we present the distance and time Nyquist frequencies for the absolute and relative
examples we used earlier.
Table 14.3: Nyquist spatial and temporal frequencies.
Absolute Relative
ifk.sh dx=1 dt=1 dx=50 dt=0.004
Nyquist kN = 0.5 fN = 0.5 Hz kN = 0.01 fN = 125 Hz
Figure 14.7 is a stretched version of Figure 14.1, right. Here, the horizontal
annotations are easier to see.

Figure 14.7: An f-k plot of the input data, made using absolute dx and dt values.
Figure 14.8 is a stretched version of Figure 14.3, right. Here again, the horizontal
annotations are easier to see.

14-16
 f-k Filter and Deconvolution

For both Figures 14.7 and 14.8, the maxima of the axes (negative and positive) are the
respective Nyquist frequencies, spatial and temporal.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 14.8: An f-k plot of the input data, made using relative dx and dt values.

14-17
 f-k Filter and Deconvolution

14.5 Deconvolution
“Deconvolution is a process that improves the temporal resolution of seismic data by
compressing the basic seismic wavelet (Yilmaz, 1987, Yilmaz 2001: Chapter 2).” The
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

two important parameters of deconvolution in supef (Wiener predictive error filtering)

are minlag (seconds) and maxlag (seconds). Our deconvolution script, idecon, refers to
these as prediction length and operator length (respectively). We take this terminology
from Robinson and Treitel (2000). In explaining Figure 14.9, Robinson and Treitel
(2000) write, “Empirical studies indicate that Į [prediction length] should be set roughly
equal to the lag that corresponds to the second zero crossing of the autocorrelation
function.”
Note: Operator length must be longer than prediction length.

Figure 14.9: Top: Autocorrelation of trace with reverberation. Bottom: Same trace after
deconvolution using prediction distance Į and operator length n (Robinson and Treitel,
2000, Figure 12-3). Used with permission.

14.5.1 Use Deconvolution Script idecon

For our interactive deconvolution processing example, we use Oz gather 16 from the
CSM CWP web site (Section 3.3). Below is the surange output of oz16.su.

14-18
 f-k Filter and Deconvolution

surange < oz16.su

48 traces:
tracl=(1,48) tracr=(1,48) fldr=10016 tracf=(1,48) cdp=(16,63)
cdpt=1 trid=1 nvs=1 nhs=1 duse=1
scalel=1 scalco=1 counit=1 delrt=4 muts=4
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

ns=1325 dt=4000
We want to use only 3 seconds of this 4 millisecond sample interval data set, and we
want to mute the strong refractions. We examine the data using the following line
command:
sugain < oz16.su tpow=2 | suxwigb perc=95 &
The following line command mutes the early arrivals and windows the first 3
seconds:
sumute < oz16.su key=tracl tmute=1.200,0.395 xmute=1,48 | suwind tmax=3 >
oz16m3.su
Below is the surange output of oz16m3.su.
surange < oz16m3.su
48 traces:
tracl=(1,48) tracr=(1,48) fldr=10016 tracf=(1,48) cdp=(16,63)
cdpt=1 trid=1 nvs=1 nhs=1 duse=1
scalel=1 scalco=1 counit=1 delrt=4 muts=(395,1200)
ns=750 dt=4000
As we stated in the introduction, script idecon uses programs supef and suacor.
Program supef does the deconvolution. Script idecon uses suacor to make plots of each
trace’s autocorrelation, an important deconvolution analysis tool. Below is the parameter
list from the supef selfdoc:
SUPEF - Wiener predictive error filtering

supef <stdin >stdout [optional parameters]

Required parameters:
dt is mandatory if not set in header

Optional parameters:
minlag=dt first lag of prediction filter (sec)
maxlag=last lag default is (tmax-tmin)/20
pnoise=0.001 relative additive noise level
mincorr=tmin start of autocorrelation window (sec)
maxcorr=tmax end of autocorrelation window (sec)
showwiener=0 =1 to show Wiener filter on each trace
mix=1,... array of weights (floats) for moving
average of the autocorrelations
Notice that supef has windowing parameters for the autocorrelation (parameters mincorr
and maxcorr).
Below is the suacor selfdoc:
SUACOR - auto-correlation

suacor <stdin >stdout [optional parms]

Optional Parameters:
ntout=101 odd number of time samples output
norm=1 if non-zero, normalize maximum absolute output to 1
sym=1 if non-zero, produce a symmetric output from
lag -(ntout-1)/2 to lag +(ntout-1)/2
Notice that suacor does not have any windowing parameters.

14-19
 f-k Filter and Deconvolution

In idecon, if the user decides to window the autocorrelation prior to applying

deconvolution, idecon windows the traces that go into suacor (that go into the
autocorrelation plot).
There are only two values that must be supplied before using idecon: the name of the
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

seismic file and the percent white noise for supef (supef parameter pnoise). All other
parameters are input in response to questions that appear on the screen. Below is the
screen dialog with our input highlighted. Line numbers are added for discussion. We
input values on lines 8, 14, 18, 22, 37, 40, 42, 50, 52, 56, 60, 74, 77, 79, 87, 89, and 93.
1
2 *** DECONVOLUTION TEST ***
3
4 ENVIRONMENT QUESTIONS ...
5
6 Supply gain power value for t^(power)
7 For no gain, supply 0
8 2
9
10 Supply "perc" value for plots
11 Typical values are 100, 98, 95, 90, 85
12 If you need a perc value less than 70,
13 your data are probably poorly gained.
14 98
15
16 Supply "key" for plot x-axis annotation
17 Typical values are: tracl tracf offset
18 tracl
19
20 For wiggle traces, enter w
21 For image display, enter i
22 w
23
24 ITERATION QUESTIONS ...
25
26 oz16m3.su:
27
28 ----------------------------------------
29 Sample interval is .004000 seconds
30 Data last time is 3.000000 seconds
31 ----------------------------------------
32
33 Autocorrelation window start time is .004000
34 Autocorrelation window last time is 3.000000
35
36 Change autocorrelation analysis window ? (y/n)
37 y
38
39 Supply first time (seconds)
40 1.
41 Supply last time (seconds)
42 2.8
43
44 oz16m3.su:
45
46 Autocorrelation first time is 1.
47 Autocorrelation last time is 2.8
48
49 Enter prediction length (seconds)
50 .030
51 Enter operator length (seconds)
52 0.175

14-20
 f-k Filter and Deconvolution

53
54 Enter 1 for more decon testing
55 Enter 2 for EXIT
56 1
57
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

58 Enter R to re-process deconvolved data

59 Enter S to start over
60 r
61 -> Using modified data
62
63 oz16m3.su:
64
65 ----------------------------------------
66 Sample interval is .004000 seconds
67 Data last time is 3.000000 seconds
68 ----------------------------------------
69
70 Autocorrelation window start time is 1.
71 Autocorrelation window last time is 2.8
72
73 Change autocorrelation analysis window ? (y/n)
74 y
75
76 Supply first time (seconds)
77 .004
78 Supply last time (seconds)
79 3.000
80
81 oz16m3.su:
82
83 Autocorrelation first time is .004
84 Autocorrelation last time is 3.000
85
86 Enter prediction length (seconds)
87 .004
88 Enter operator length (seconds)
89 .040
90
91 Enter 1 for more decon testing
92 Enter 2 for EXIT
93 2
94
95 Output seismic file is ==> idecon.su
96 Output log file is ==> idecon.txt
97
98 press return key to exit
99
In the beginning, idecon asks for “environment” values: a gain value, a perc plot
value, a key for labeling the suxwigb x-axis, and whether the user wants to see wiggle
plots or image plots. The next questions are asked with each loop through idecon.
Dialog lines 29 and 30 are printed with each loop to remind the user of the sample
interval and the total time of the data. By knowing the sample interval, the user knows
which value to supply for spiking deconvolution. Initially, these two values are used to
set the autocorrelation window (dialog lines 33-34). Before the first autocorrelation plot
is made, the user is given the chance to change start and stop time of the autocorrelation
window (line 36). In this case, the user changes the values (dialog lines 39-42). The user
is immediately given a confirmation message of the autocorrelation window values. At
this time, the right side of Figure 14.10 is made.

14-21
 f-k Filter and Deconvolution
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 14.10: Left: Seismic data for deconvolution. Right: Autocorrelation of the
seismic traces. In this first decon test, the autocorrelation plot was made using seismic
traces windowed (suwind) from 1 second to 2.8 seconds. (See dialog lines 36-42).

Figure 14.11: Left: Seismic data after deconvolution. Right: Autocorrelation of the
windowed portion of the deconvolved seismic traces.
After the user supplies a prediction length and an operator length (dialog lines 49-52),
the plots of Figure 14.11 are made (a gapped deconvolution test).

14-22
 f-k Filter and Deconvolution

For the second test, the user decides to deconvolve the output of the first test (instead
of using the original input (dialog lines 58-61). The user is reminded of the seismic file’s
sample interval and total time (dialog lines 66-67) as well as the current settings for the
autocorrelation window (dialog lines 70-71).
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Again, the user decides to change the autocorrelation window, supplying values that
correspond to the entire trace (dialog lines 73-79).
For the new test, the user supplies a prediction length of one sample and an operator
length (dialog lines 86-89), and the plots of Figure 14.12 are made for a “spiking”
deconvolution. The autocorrelation plot of Figure 14.12 is made from the entire trace.
The seismic plots show the prediction length and the operator length, while the
autocorrelation plots show the extent of the autocorrelation windows.

Figure 14.12: Left: Seismic data after spiking deconvolution. Right: Autocorrelation
of the deconvolved seismic traces (using the entire times).
Below are the contents of idecon.txt, the log file. Line numbers are added for
discussion. Line 2 includes the percent white noise and the gain (t^(power)) values.
1 *** DECONVOLUTION TEST ***
2 Input file = oz16m3.su
3 Percent white noise = 0.001 Gain = 2
4 Sample interval is .004000 seconds
5 Data last time is 3.000000 seconds
6 --------------------------
7 -> Original data
8 Autocorrelation first time is 1.
9 Autocorrelation last time is 2.8
10 Prediction length = .030
11 Operator length = 0.175
12 -> Using modified data
13 Autocorrelation first time is .004
14 Autocorrelation last time is 3.000
15 Prediction length = .004

14-23
 f-k Filter and Deconvolution

16 Operator length = .040

14.5.2 Script idecon.scr

We usually start interactive scripts by running a .scr script. Here is idecon.scr:
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

1 #! /bin/sh
2 # File: idecon.scr
3 # Run this script to start script idecon.sh
4
5 xterm -geom 60x18+10+544 -e idecon.sh
6
Line 1 invokes the shell. Line 5 opens a dialog window 80 characters wide by 18
characters high. Line 5 also places the dialog window 10 pixels from the left side of the
viewing area and 544 pixels from the top of the viewing area. Line 5 also starts idecon.sh,
the processing script, within that window.

14.5.3 Script idecon.sh

We start interactive scripts by running a .scr script. Here is idecon.sh.
1 #! /bin/sh
2 # File: idecon.sh
3 # Run script idecon.scr to start this script
4 # You can window the seismic data for autocorrelation
5 # analysis. For predictive deconvolution, based on
6 # the autocorrelation display, use a prediction length
7 # equal to the time from zero to the second zero crossing.
8 # Example:
9 # Prediction length = .030 seconds
10 # Operator length = .250 seconds
11
12 # Set messages on
13 ##set -x
14
15 #================================================
16 # User-supplied values
17 #------------------------------------------------
18
19 # Input seismic file
20 indata=oz16m3.su
21
22 # percent white noise [supef default=0.001]
23 wnoise=0.001
24
25 #================================================
26
27 # DOCUMENT TERMS:
28 #
29 # PL = Prediction length
30 # OL = Operator length
31 #
32 # |
33 # **
34 # | * **
35 # | * * * **
36 # | * * * * * **
37 # ------*------*---- *----*-- *--*--*--*
38 # | * *| * * ** |**
39 # | * * | ** |
40 # | ** | |
41 # | | |

14-24
 f-k Filter and Deconvolution

42 # | == PL === | ======== OL ======== |

43 #
44 # Input wavelet
45 #
46 #
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

47 # |
48 # **
49 # | *
50 # | *
51 # | *
52 # ------*------************************--*
53 # | * * **
54 # | * *
55 # | **
56 # |
57 #
58 # Output wavelet
59 #
60 #
61 # REFERENCE: Robinson, E.A., Treitel, S., 2000, Geophysical
62 # Signal Analysis, Society of Exploration
63 # Geophysicists, Tulsa, p. 278-279.
64 #
65 #------------------------------------------------
66 # Describe temporary files
67 #------------------------------------------------
68
69 # tmp0 = input seismic file, after gain
70 # tmp1 = input to deconvolution:
71 # tmp0 (reproc=s) or tmp3 (reproc=r)
72 # tmp2 = input seismic windowed for autocorrelation
73 # tmp3 = output of the prediction error filter
74 # tmp4 = output of the prediction error filter
75 # windowed for autocorrelation
76 # tmp5 = ASCII log file of processing parameters
77 # tmp6 = ASCII file to reduce screen display of "zap"
78
79 #------------------------------------------------
80
81 echo " "
82 echo " *** DECONVOLUTION TEST ***"
83 echo " "
84 echo " ENVIRONMENT QUESTIONS ..."
85
86 # Remove temporary files
87 rm -f tmp*
88
89 #------------------------------------------------
90 # Supply gain value to t^(power)
91 # Supply perc value for plots
92 # Supply key value for plots
93
94 echo " "
95 echo "Supply gain power value for t^(power)"
96 echo "For no gain, supply 0"
97 > /dev/tty
98 read tpow
99
100 echo " "
101 echo "Supply \"perc\" value for plots"
102 echo " Typical values are 100, 98, 95, 90, 85"
103 echo " If you need a perc value less than 70,"
104 echo " your data are probably poorly gained."

14-25
 f-k Filter and Deconvolution

105 > /dev/tty

106 read myperc
107
108 echo " "
109 echo "Supply \"key\" for plot x-axis annotation"
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

110 echo " Typical values are: tracl tracf offset"

111 > /dev/tty
112 read mykey
113
114 #------------------------------------------------
115 # Ask: wiggle traces (xwigb) or image (ximage)
116
117 echo " "
118 echo "For wiggle traces, enter w"
119 echo "For image display, enter i"
120 > /dev/tty
121 read display
122
123 case $display in
124 i)
125 wiggle=1
126 ;;
127 *)
128 wiggle=0
129 ;;
130 esac
131
132 #------------------------------------------------
133 # Apply gain, copy input data to temporary file
134
135 sugain < $indata tpow=$tpow > tmp0
136
137 #------------------------------------------------
138 # From data: get number of samples, sample interval
139 # Compute time length of traces
140
141 ns=`sugethw ns < tmp0 | sed 1q | sed 's/.*ns=//'`
142 dt=`sugethw dt < tmp0 | sed 1q | sed 's/.*dt=//'`
143
144 # Convert dt key value from microsec to sec
145 tsamp=`bc -l << -END
146 scale=6
147 $dt / 1000000
148 END`
149
150 # Compute last time (seconds)
151 tend=`bc -l << -END
152 $ns * $tsamp
153 END`
154
155 #------------------------------------------------
156 # Write to log file
157
158 echo " *** DECONVOLUTION TEST ***" > tmp5
159 echo " Input file = $indata" >> tmp5
160 echo " Percent white noise = $wnoise Gain = $tpow" >> tmp5
161 echo " Sample interval is $tsamp seconds" >> tmp5
162 echo " Data last time is $tend seconds" >> tmp5
163 echo " --------------------------" >> tmp5
164
165 #------------------------------------------------
166
167 # initial setting for autocorrelation analysis window start

14-26
 f-k Filter and Deconvolution

168 taa=$tsamp
169 # initial setting for autocorrelation analysis window end
170 tzz=$tend
171
172 #------------------------------------------------
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

173
174 new=true # true = 1st time through loop
175 finish=false # true = exit decon script
176
177 #------------------------------------------------
178 # Begin testing loop
179 #------------------------------------------------
180
181 echo " "
182 echo " ITERATION QUESTIONS ..."
183
184 while [ $finish = false ]
185 do
186
187 if [ $new = true ] ; then
188
189 cp tmp0 tmp1
190 echo " -> Original data" >> tmp5
191
192 else
193
194 echo " "
195 echo "Enter R to re-process deconvolved data"
196 echo "Enter S to start over"
197 > /dev/tty
198 read reproc
199
200 case $reproc in
201 [sS])
202 cp tmp0 tmp1
203 echo " -> Using original data"
204 echo " -> Using original data" >> tmp5
205 ;;
206 [rR])
207 cp tmp3 tmp1
208 echo " -> Using modified data"
209 echo " -> Using modified data" >> tmp5
210 ;;
211 esac
212
213 fi
214
215 rm -f tmp3 tmp4
216
217 #------------------------------------------------
218 # Plot seismic
219
220 if [ $wiggle -eq 0 ] ; then
221 suxwigb < tmp1 perc=$myperc xbox=10 ybox=10 wbox=300 hbox=500 \
222 label1=" Time (s)" label2=$mykey key=$mykey \
223 windowtitle="Test input" \
224 title="Input to deconvolution" verbose=0 &
225 else
226 suximage < tmp1 perc=$myperc xbox=10 ybox=10 wbox=300 hbox=500 \
227 label1=" Time (s)" \
228 windowtitle="Test input" \
229 title="Input to deconvolution" verbose=0 &
230 fi

14-27
 f-k Filter and Deconvolution

231
232 #------------------------------------------------
233 # Tell user current autocorrelation window settings
234
235 echo " "
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

236 echo " $indata:"

237 echo " "
238 echo " ----------------------------------------"
239 echo " Sample interval is $tsamp seconds"
240 echo " Data last time is $tend seconds"
241 echo " ----------------------------------------"
242 echo " "
243 echo " Autocorrelation window start time is $taa"
244 echo " Autocorrelation window last time is $tzz"
245
246 #------------------------------------------------
247 # Autocorrelation analysis window
248
249 echo " "
250 echo "Change autocorrelation analysis window ? (y/n)"
251 > /dev/tty
252 read windowC
253
254 case $windowC in
255 [yY])
256 wndwC=0
257 ;;
258 *)
259 wndwC=1
260 ;;
261 esac
262
263 if [ $wndwC -eq 0 ] ; then
264 echo " "
265 echo "Supply first time (seconds)"
266 > /dev/tty
267 read taa
268 echo "Supply last time (seconds)"
269 > /dev/tty
270 read tzz
271 fi
272
273 echo " "
274 echo " $indata:"
275 echo " "
276 echo " Autocorrelation first time is $taa"
277 echo " Autocorrelation first time is $taa" >> tmp5
278 echo " Autocorrelation last time is $tzz"
279 echo " Autocorrelation last time is $tzz" >> tmp5
280
281 #------------------------------------------------
282 # Plot autocorrelation
283
284 suwind < tmp1 tmin=$taa tmax=$tzz > tmp2
285
286 if [ $wiggle -eq 0 ] ; then
287 suacor < tmp2 ntout=101 sym=0 |
288 suxwigb perc=$myperc xbox=322 ybox=10 wbox=300 hbox=500 \
289 label1=" Time (s)" label2=$mykey key=$mykey \
290 windowtitle="Acor before decon" \
291 title="Autocorrelation" wclip=0 verbose=0 &
292 else
293 suacor < tmp2 ntout=101 sym=0 |

14-28
 f-k Filter and Deconvolution

294 suximage perc=$myperc xbox=322 ybox=10 wbox=300 hbox=500 \

295 label1=" Time (s)" \
296 windowtitle="Acor before decon" \
297 title="Autocorrelation" wclip=0 verbose=0 &
298 fi
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

299
300 #------------------------------------------------
301 # Prediction length & Operator length
302
303 echo " "
304 echo "Enter prediction length (seconds)"
305 > /dev/tty
306 read minlag
307 echo "Enter operator length (seconds)"
308 > /dev/tty
309 read maxlag
310
311 echo " Prediction length = $minlag" >> tmp5
312 echo " Operator length = $maxlag" >> tmp5
313
314 #------------------------------------------------
315 # Deconvolution
316
317 supef < tmp1 minlag=$minlag maxlag=$maxlag \
318 mincorr=$taa maxcorr=$tzz pnoise=$wnoise > tmp3
319
320 #------------------------------------------------
321 # After deconvolution -- plot data
322
323 if [ $wiggle -eq 0 ] ; then
324 suxwigb < tmp3 perc=$myperc xbox=634 ybox=10 wbox=300 hbox=500 \
325 label1=" Time (s)" label2=$mykey key=$mykey \
326 title="Pred length=$minlag, Oper length=$maxlag" \
327 windowtitle="Deconvolution" verbose=0 &
328 else
329 suximage < tmp3 perc=$myperc xbox=634 ybox=10 wbox=300 hbox=500 \
330 label1=" Time (s)" \
331 title="Pred length=$minlag, Oper length=$maxlag" \
332 windowtitle="Deconvolution" verbose=0 &
333 fi
334
335 #------------------------------------------------
336 # After deconvolution -- plot autocorrelation
337
338 suwind < tmp3 tmin=$taa tmax=$tzz > tmp4
339
340 if [ $wiggle -eq 0 ] ; then
341 suacor <tmp4 ntout=101 sym=0 |
342 suxwigb perc=$myperc xbox=946 ybox=10 wbox=300 hbox=500 \
343 label1=" Time (s)" label2=$mykey key=$mykey \
344 title="Acor: first time = $taa, last time = $tzz" \
345 windowtitle="Acor after decon" verbose=0 &
346 else
347 suacor < tmp4 ntout=101 sym=0 |
348 suximage perc=$myperc xbox=946 ybox=10 wbox=300 hbox=500 \
349 label1=" Time (s)" \
350 title="Acor: first time = $taa, last time = $tzz" \
351 windowtitle="Acor after decon" verbose=0 &
352 fi
353
354 #------------------------------------------------
355 # Do more decon or exit
356

14-29
 f-k Filter and Deconvolution

357 echo " "

358 echo "Enter 1 for more decon testing"
359 echo "Enter 2 for EXIT"
360 > /dev/tty
361 read selection
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

362
363 case $selection in
364 1)
365 finish=false
366 ;;
367 2)
368 cp tmp3 idecon.su
369 cp tmp5 idecon.txt
370 echo " "
371 echo "Output seismic file is ==> idecon.su"
372 echo " Output log file is ==> idecon.txt"
373 pause exit
374 zap xwigb > tmp6 # decrease screen messages
375 zap ximage > tmp6 # decrease screen messages
376 finish=true
377 ;;
378 esac
379
380 new=false
381
382 done
383
384 #------------------------------------------------
385 # Exit
386 #------------------------------------------------
387
388 # Remove temporary files
389 rm -f tmp*
390
391 # Exit politely from shell
392 exit
393
Script idecon.sh has the following major sections:
• Lines 19-23 are user-supplied values: the input file and a percent white noise
value.
• Lines 27-64 describe prediction length and operator length and provide a
reference. Also, lines 66-77 describe the temporary files.
• Lines 81-182 operate before the deconvolution test loop.
o Line 87 removes old temporary files.
o Lines 95-98 get the user’s gain value.
o Lines 101-106 get the user’s perc value.
o Lines 109-112 get the users’ key plot value.
o Lines 118-130 get the user’s preference for wiggle or image plots.
o Line 135 applies gain to the input data and holds the output in tmp0.
o Lines 141-153 compute the sample interval (seconds) and the time length of
the traces (seconds).
o Lines 158-163 write information to the log file.

14-30
 f-k Filter and Deconvolution

o Lines 167-170 create initial autocorrelation window analysis start and stop
times – the values computed in lines 141-153.
o Line 174 sets a flag that means this is the first time through the test loop.
o Line 175 sets a flag that means the user is not finished testing.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

• Lines 185-382 are the test loop.

o Lines 187-190: If this is the first time the user is in the test loop (new=true),
the contents of tmp0 are copied to tmp1. File tmp1 is the file that is displayed
(lines 220-230) and windowed for autocorrelation analysis and display (lines
284-298). The windowed contents of tmp1 are output to tmp2. File tmp2 is
converted to autocorrelation traces and displayed (lines 286-291 or lines 293-
297).
o Lines 192-211: If this is not the first test (new=false), the user chooses
whether to re-process the deconvolved data (tmp3) or process the original data
(tmp0).
o Line 215 removes the two files that are created after deconvolution. File tmp3
holds the deconvolution output (made by supef, lines 317-318). Files tmp4
holds the deconvolved and time-windowed data (line 338) that are converted
to autocorrelation traces and displayed (lines 340-345 or lines 346-351).
o Lines 233-279 inform the user of the current autocorrelation window limits
(lines 236-244), allow the user to set new limits (lines 250-271) , and again
inform the user of the autocorrelation window limits (lines 276-279).
o After the test seismic data are displayed and the autocorrelation traces are
displayed, the user is asked for prediction length and operator length (lines
304-309). The user-supplied values are written to the log file (lines 311-312).
o After the deconvolution results and the post-deconvolution autocorrelation
traces are displayed, the user chooses whether to run a new test or exit the
script (lines 358-361). If the user chooses to exit (lines 367-377), the last
seismic deconvolution output data are written to a disk file (line 368), the
temporary log file is written to a disk file (line 369), screen messages inform
the user of the names of the disk files (lines 371-372), the displays are closed
(lines 374-375), and the looping flag is set to end the loop (finish=true).
o Line 380 sets a flag, new=false. This means the user will always have to
choose whether the next test re-processes the deconvolved data or uses the
original data (lines 192-211). Line 380 is at the bottom of the loop, but could
be placed just after the new test, on line 214.
We encourage you to try these tools on the Nankai and Taiwan data sets.

14-31
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

This page has been intentionally left blank

 Taiwan: Sort, Velocity Analysis, NMO, Stack

15. Taiwan: Sort, Velocity Analysis, NMO, Stack

15.1 Introduction
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Returning to the Taiwan data, we now have 5 seconds of data, we zeroed the bad
hydrophones, applied spherical divergence correction, and band-pass filtered the shot
gathers. The result is Tshot5kgf.su. We will now
• sort these shot gathers to CMP order,
• analyze the CMPs for stacking velocities,
• use the stacking velocities to apply NMO, and
• stack the CMPs.

15.2 Sort
Sorting to CMP order is a one-line command:
susort < Tdata/Tshot5kgf.su cdp offset > Tdata/Tcmp5kgf.su
We send the sorted data to “Tdata” since this file is just as large as the file of windowed
(0-5 seconds) shot gathers: 133 Mbytes.

Figure 15.1: Fold values for data set Tcmp5kgf.su.

Below is the surange output of Tcmp5kgf.su.
surange < Tcmp5kgf.su
25344 traces:

15-1
 Taiwan: Sort, Velocity Analysis, NMO, Stack

tracl=(1,25344) tracr=(1,25344) fldr=(800,975) tracf=(1,144) ep=(740,915)

cdp=(4027,4816) cdpt=(1,144) trid=1 nhs=1 offset=(-3663,-88)
sdepth=80000 swdep=(2120000,3540000) scalel=-10000 scalco=1 sx=(-4773100,-
3965250)
gx=(-4764600,-3600343) gy=(766,94805) counit=3 tstat=12 ns=1251
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

dt=4000 gain=9 afilf=160 afils=72 lcf=3

hcf=160 lcs=6 hcs=72 year=95 day=260
hour=(12,13) minute=(0,59) sec=(0,59)
We have 790 CMP gathers (4816-4027+1). Figure 15.1 shows that the full fold
gathers (at least 34 traces per gather) are approximately 4175-4700. (We describe the
program, sukeycount, which generated the data for Figure 15.1 in Appendix C.) Despite
the constant number of traces per shot, CMP fold varies (Figure 15.1) because ship speed
and streamer position varied.

15.3 Velocity Analysis

Below is the user area of iva.sh changed for the Taiwan data. To make reasonably
detailed velocity analysis at a regular interval along the line, we choose to analyze every
25th CMP starting with CMP 4175 (lines 13-18). We set the perc value to 90 (line 31), an
accommodation for real data. Semblance values go from 1000 m/s (line 38) to 5000 m/s
(nvs*dvs + fvs). CVS panels range from 1000 m/s to 5000 m/s (lines 43-46).
8 #================================================
9 # USER AREA -- SUPPLY VALUES
10 #------------------------------------------------
11 # CMPs for analysis
12
13 cmp1=4175 cmp2=4200 cmp3=4225 cmp4=4250
14 cmp5=4275 cmp6=4300 cmp7=4325 cmp8=4350
15 cmp9=4375 cmp10=4400 cmp11=4425 cmp12=4450
16 cmp13=4475 cmp14=4500 cmp15=4525 cmp16=4550
17 cmp17=4575 cmp18=4600 cmp19=4625 cmp20=4650
18 cmp21=4675 cmp22=4700
19
20 numCMPs=22
21
22 #------------------------------------------------
23 # File names
24
25 indata=Tdata/Tcmp5kgf.su # SU format
26 outpicks=Tvpick.txt # ASCII file
27
28 #------------------------------------------------
29 # display choices
30
31 myperc=90 # perc value for plot
32 plottype=1 # 0 = wiggle plot, 1 = image plot
33
34 #------------------------------------------------
35 # Processing variables
36
37 # Semblance variables
38 nvs=160 # number of velocities
39 dvs=25 # velocity intervals
40 fvs=1000 # first velocity
41
42 # CVS variables
43 fc=1000 # first CVS velocity
44 lc=5000 # last CVS velocity
45 nc=10 # number of CVS velocities (panels)

15-2
 Taiwan: Sort, Velocity Analysis, NMO, Stack

46 XX=11 # ODD number of CMPs to stack into central CVS

47
48 #================================================
We make one more accommodation for real data. When we use suximage following
suvelan, we set perc=97 (highlighted below). These two lines occur twice, once for
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

plotting the velan with picks overlain and once for plotting the original velan. See lines
305 and 316 in Section 7.6.6.C.
suvelan < panel.$picknow nv=$nvs dv=$dvs fv=$fvs |
suximage xbox=10 ybox=10 wbox=300 hbox=450 perc=97 \
By changing perc from 99 (in Section 7.6.6.C) to 97, we “increase the volume” of the
semblance values.
Below are two of the semblance plots with picks overlain. Our philosophy was to
pick generally increasing velocities. Better picks might be made by following subtle
events. We encourage you to experiment.

Figure 15.2: Semblance plots. Left: CMP 4225. Right: CMP 4575.

15-3
 Taiwan: Sort, Velocity Analysis, NMO, Stack

Below is the file of t-v picks. For the sake of presentation, we made line 1 into two
lines. Line 1 as presented here has no spaces at the end of the first part and no spaces at
the beginning of the second part, making this a usable file. Remember to remove line 2
before using this file in sunmo.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

1 cdp=4175,4200,4225,4250,4275,4300,4325,4350,4375,4400,4425,\
4450,4475,4500,4525,4550,4575,4600,4625,4650,4675,4700 \
2 #=1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22 \
3 tnmo=0.0527704,0.527704,2.26913,3.219,3.7467,4.96042 \
4 vnmo=1485.83,1574.17,2325,2634.17,3252.5,4224.17 \
5 tnmo=0.0659631,1.0686,3.33773,5 \
6 vnmo=1463.75,1706.67,3053.75,3826.67 \
7 tnmo=0.0527704,0.844327,2.00528,3.37731,4.93404 \
8 vnmo=1507.92,1640.42,2325,3031.67,3426.41 \
9 tnmo=0.0395778,0.897098,2.41425,3.49604,4.97361 \
10 vnmo=1530,1640.42,2347.08,3009.58,4069.58 \
11 tnmo=0.0527704,0.620053,2.71768,3.64116,4.98681 \
12 vnmo=1463.75,1596.25,2479.58,2832.92,3583.75 \
13 tnmo=0.0395778,0.633245,2.62533,3.48285,4.98681 \
14 vnmo=1463.75,1596.25,2656.25,3208.33,3892.92 \
15 tnmo=0.0527704,0.633245,3.7467,4.98681 \
16 vnmo=1441.67,1552.08,3451.25,3848.75 \
17 tnmo=0.0527704,0.699208,2.8496,4.98681 \
18 vnmo=1507.92,1662.5,2567.92,3053.75 \
19 tnmo=0.0263852,1.04222,3.85224,4.97361 \
20 vnmo=1441.67,1772.92,3164.17,3716.25 \
21 tnmo=0.0527704,1.22691,3.10026,4.12929,5 \
22 vnmo=1463.75,1728.75,2590,3539.58,3804.58 \
23 tnmo=0.0527704,1.18734,2.4934,3.66755,5 \
24 vnmo=1463.75,1684.58,2413.33,2899.17,3252.5 \
25 tnmo=0.0527704,1.10818,2.94195,3.69393,4.97361 \
26 vnmo=1485.83,1706.67,2413.33,2921.25,3296.67 \
27 tnmo=0.0527704,0.91029,3.68074,4.98681 \
28 vnmo=1507.92,1772.92,3495.42,4091.67 \
29 tnmo=0.0659631,1.21372,3.69393,4.97361 \
30 vnmo=1530,1596.25,2965.42,3738.33 \
31 tnmo=0.0527704,1.21372,3.44327,5.01319 \
32 vnmo=1485.83,1662.5,2766.67,3694.17 \
33 tnmo=0.0395778,1.26649,3.79947,4.97361 \
34 vnmo=1463.75,1640.42,2877.08,3097.92 \
35 tnmo=0.0527704,2.24274,4.1161,4.97361 \
36 vnmo=1463.75,1883.33,2457.5,2943.33 \
37 tnmo=0.0395778,1.93931,3.25858,5.01319 \
38 vnmo=1441.67,1927.5,2457.5,3451.25 \
39 tnmo=0.0659631,1.27968,3.25858,4.97361 \
40 vnmo=1485.83,1618.33,2810.83,3826.67 \
41 tnmo=0.0527704,1.54354,3.13984,4.97361 \
42 vnmo=1485.83,1861.25,2766.67,3318.75 \
43 tnmo=0.0527704,1.41161,3.50923,4.98681 \
44 vnmo=1507.92,1905.42,2943.33,3340.83 \
45 tnmo=0.0527704,1.20053,2.3219,3.0343,4.98681 \
46 vnmo=1485.83,1839.17,2722.5,3031.67,3407.08 \

15.4 NMO and Stack

Below is file Tnmo.sh. The output stack file size is 4.14 Mbytes.
1 #! /bin/sh
2 # File: Tnmo.sh
3 # Apply NMO & stack
4 # Input: line of CMPs
5 # Output: stacked seismic data
6

15-4
 Taiwan: Sort, Velocity Analysis, NMO, Stack

7 # Set messages on
8 set -x
9
10 # User values
11 indata=Tdata/Tcmp5kgf.su
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

12 outdata=Tstack.su
13 myperc=95
14
15 # NMO
16 sunmo < $indata \
17 cdp=4175,4200,4225,4250,4275,4300,4325,4350,4375,4400,4425,
4450,4475,4500,4525,4550,4575,4600,4625,4650,4675,4700 \
18 tnmo=0.0527704,0.527704,2.26913,3.219,3.7467,4.96042 \
19 vnmo=1485.83,1574.17,2325,2634.17,3252.5,4224.17 \
20 tnmo=0.0659631,1.0686,3.33773,5 \
21 vnmo=1463.75,1706.67,3053.75,3826.67 \
22 tnmo=0.0527704,0.844327,2.00528,3.37731,4.93404 \
23 vnmo=1507.92,1640.42,2325,3031.67,3426.41 \
24 tnmo=0.0395778,0.897098,2.41425,3.49604,4.97361 \
25 vnmo=1530,1640.42,2347.08,3009.58,4069.58 \
26 tnmo=0.0527704,0.620053,2.71768,3.64116,4.98681 \
27 vnmo=1463.75,1596.25,2479.58,2832.92,3583.75 \
28 tnmo=0.0395778,0.633245,2.62533,3.48285,4.98681 \
29 vnmo=1463.75,1596.25,2656.25,3208.33,3892.92 \
30 tnmo=0.0527704,0.633245,3.7467,4.98681 \
31 vnmo=1441.67,1552.08,3451.25,3848.75 \
32 tnmo=0.0527704,0.699208,2.8496,4.98681 \
33 vnmo=1507.92,1662.5,2567.92,3053.75 \
34 tnmo=0.0263852,1.04222,3.85224,4.97361 \
35 vnmo=1441.67,1772.92,3164.17,3716.25 \
36 tnmo=0.0527704,1.22691,3.10026,4.12929,5 \
37 vnmo=1463.75,1728.75,2590,3539.58,3804.58 \
38 tnmo=0.0527704,1.18734,2.4934,3.66755,5 \
39 vnmo=1463.75,1684.58,2413.33,2899.17,3252.5 \
40 tnmo=0.0527704,1.10818,2.94195,3.69393,4.97361 \
41 vnmo=1485.83,1706.67,2413.33,2921.25,3296.67 \
42 tnmo=0.0527704,0.91029,3.68074,4.98681 \
43 vnmo=1507.92,1772.92,3495.42,4091.67 \
44 tnmo=0.0659631,1.21372,3.69393,4.97361 \
45 vnmo=1530,1596.25,2965.42,3738.33 \
46 tnmo=0.0527704,1.21372,3.44327,5.01319 \
47 vnmo=1485.83,1662.5,2766.67,3694.17 \
48 tnmo=0.0395778,1.26649,3.79947,4.97361 \
49 vnmo=1463.75,1640.42,2877.08,3097.92 \
50 tnmo=0.0527704,2.24274,4.1161,4.97361 \
51 vnmo=1463.75,1883.33,2457.5,2943.33 \
52 tnmo=0.0395778,1.93931,3.25858,5.01319 \
53 vnmo=1441.67,1927.5,2457.5,3451.25 \
54 tnmo=0.0659631,1.27968,3.25858,4.97361 \
55 vnmo=1485.83,1618.33,2810.83,3826.67 \
56 tnmo=0.0527704,1.54354,3.13984,4.97361 \
57 vnmo=1485.83,1861.25,2766.67,3318.75 \
58 tnmo=0.0527704,1.41161,3.50923,4.98681 \
59 vnmo=1507.92,1905.42,2943.33,3340.83 \
60 tnmo=0.0527704,1.20053,2.3219,3.0343,4.98681 \
61 vnmo=1485.83,1839.17,2722.5,3031.67,3407.08 |
62
63 # Stack
64 sustack > $outdata
65
66 suximage < $outdata perc=$myperc label2=cdp key=cdp \
67 title="$outdata perc=$myperc" &
68

15-5
 Taiwan: Sort, Velocity Analysis, NMO, Stack

69 # Exit politely from shell

70 exit
71
Script iva puts a continuation mark (\) at the end of all lines. At the end of line 61, we
replaced the continuation mark with a pipe (|). After NMO, the data are piped to sustack
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

(line 64).
Lines 66-67 make a plot using suximage. Remember, suximage reads the first key
value (here it is cdp) and increments by “1” (see Section 7.7). If key values do not
increment by “1” (for example, due to missing CMPs), the x-axis annotation will be
wrong.
Figure 15.3 shows our stack of the Taiwan data.

Figure 15.3: Stack of Taiwan data.

15-6
 Taiwan: Migration

16. Taiwan: Migration

16.1 Constant Velocity Stolt Migration with Tmigcvp.sh
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Below is the user area of Tmigcvp.sh for the Taiwan data.

• Remember, cdpmin and cdpmax refer to the range of CMPs that will have DMO
applied to them. Previously, we decided that we will always let sustolt apply
DMO to all stack traces (Section 9.2.1).
• The CDP bin distance, provided to us with the data files, is 16.667 m (parameter
dxcdp, line 22).
• Because we do not know rms velocities for these data, we set vscale=1.0 (line25).
We process for image, not for velocities.
• Our first velocity is 1000 m/s and our last is 4000 m/s. Our velocity increment is
100 m/s. Here, we choose a large velocity range because we think the migration
changes will be subtle.
11 #================================================
12 # USER AREA -- SUPPLY VALUES
13 #------------------------------------------------
14
15 # Seismic files
16 indata=Tdata/Tstack.su # SU format
17 outdata=Tdata/Tmigcvp.su # migration Constant Velocity Panels
18
19 # Migration variables
20 cdpmin=4027 # Start CDP value
21 cdpmax=4816 # End CDP value
22 dxcdp=16.667 # distance between adjacent CDP bins (m)
23 smig=1.0 # stretch factor (0.6 typical if vrms increasing)
24 # [the "W" factor] (Default=1.0)
25 vscale=1.0 # scale factor to apply to velocities (Default=1.0)
26 lstaper=20 # length of side tapers (traces) (Default=0)
27 lbtaper=100 # length of bottom taper (samples) (Default=0)
28
29 # Velocity panel variables
30 firstv=1000 # first velocity value
31 lastv=4000 # last velocity value
32 increment=100 # velocity increment
33
34 numVtest=100 # use to limit number of velocity panels
35 # otherwise, use very large value (100)
36
37 #================================================
The output of Tmigcvp.sh is 31 constant velocity migration panels, 128 Mbytes.

16.2 Migration Movie

Below is script Tmigmovie.sh for the Taiwan data. We set loop=1 (line 12) to run the
movie forward continuously. Below, fframe=1000 and dframe=100 because we re-ran
the migration (Tmigcvp.sh) for velocities 1000-4000 m/s at 100 m/s increment.
1 #! /bin/sh
2 # File: Tmigmovie.sh
3 # Run a "movie" of the migration panels
4 # Plot "title" shows panel velocity
5 # Enter "xmovie" for mouse and keyboard options

16-1
 Taiwan: Migration

6
7 set -x
8
9 indata=Tdata/Tmigcvp.su
10 perc=95
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

11
12 loop=1 # 1 = run panels forward continuously
13 # 2 = run panels back and forth continuously
14 # 0 = load all panels then stop
15
16 n1=1251 # number of time samples
17 d1=0.004 # time sample interval
18 n2=790 # number of traces per panel
19 d2=1 # trace spacing
20
21 width=550 # width of window
22 height=700 # height of window
23
24 fframe=1000 # velocity of first panel for title annotation
25 dframe=100 # panel velocity increment for title annotation
26
27 suxmovie < $indata perc=$perc loop=$loop \
28 n1=$n1 d1=$d1 n2=$n2 d2=$d2 \
29 width=$width height=$height \
30 fframe=$fframe dframe=$dframe \
31 title="Velocity %g" &
32
33 exit
34
Below is the surange output of Tmigcvp.su. Several facts from this output are useful
for supplying values to Tmigmovie.sh. On line 21, you have to supply the number of time
samples – the value of key ns. On line 17, you have to supply the time sample interval –
the value of key dt.
surange < Tmigcvp.su
24490 traces:
tracl=(1,790) tracr=(1,25344) fldr=(1,31) tracf=(1,144) ep=(740,915)
cdp=(4027,4816) cdpt=(1,144) trid=1 nvs=31 nhs=(1,48)
offset=(1000,4000) sdepth=80000 swdep=(2120000,3540000) scalel=-10000
scalco=1
sx=(-4773100,-3965250) gx=(-4764600,-3600343) gy=(2277,94805) counit=3
tstat=12
ns=1251 dt=4000 gain=9 afilf=160 afils=72
lcf=3 hcf=160 lcs=6 hcs=72 year=95
day=260 hour=(12,13) minute=(0,59) sec=(0,59)
The panel velocities are in the offset key and the number of velocities is in the nvs
key. Using these two keys, you can calculate the values for fframe and dframe (if you do
not want to check the Tmigcvp.sh values of firstv and increment).
Refer to Section 9.4 or the xmovie selfdoc for help using the mouse and keyboard
with the movie. If you switch from Movie mode to Step mode by pressing the S key, you
can step forward with the F key or backward with the B key. To restart the movie, press
the C key. To quit, press the Q key.
Viewing the movie of single velocity migrations, it was difficult for us to pick a
“best” migration. We think 2500 m/s is a reasonable image, but we are not happy with it.

16-2
 Taiwan: Migration
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Figure 16.1: Left: A portion of the 2-D line off the shelf area southwest of Taiwan. From
Berndt and Moore (1999), Figure 6a (used with permission). Right: Our 2500 m/s
constant velocity Stolt migration.
Figure 16.1 puts our migration next to an image reported by Berndt and Moore
(1999).

16.2 Discussion
While viewing the shot gathers, we did not mention that we saw sideswipe. Sideswipe
is “evidence of a structural feature which lies off to the side of a line or traverse (Sheriff,
2002).” The problem with a 2-D survey is that real geology is 3-D. A 2-D survey is

16-3
 Taiwan: Migration

processed to image information from below the line of acquisition, but acoustic energy
can reflect from side features as well as from “down.”
Figure 16.2 shows a strong sideswipe event between 3.5-4 seconds. Other shot
gathers show sideswipe at various times, generally at the same angle as in Figure 16.2
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

(for example, look carefully at shot 930).

Figure 16.2: Taiwan shot gather 950. The arrow points to sideswipe.
When sideswipe (also called “out of plane reflections”) occurs on 2-D data, we
cannot “image” it because a 2-D line does not have enough information. We usually try
various ways to eliminate sideswipe. One way is tau-p filtering. The tau-p filter is also
called slant stack processing and radon filtering (Yilmaz 1987 and Yilmaz 2001).
Although we do not discuss tau-p filtering in this Primer, SU does have suradon and
sutaup. Look at the examples in demos/Tau_P.
The paper by Berndt and Moore (1999) discusses several methods to attenuate sea-
floor multiples, one of which they applied to this data set (their Figure 6). Two processes
we recommend that you investigate are:
• prestack f-k or tau-p filtering and
• time-varying migration.

16-4
 References
Berndt, C. & Moore, G.F., 1999, Dependence of multiple-attenuation techniques on the
geologic setting: A case study from offshore Taiwan, The Leading Edge, Vol. 18,
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

No. 1, 74-80.
Cohen, J.K. & Stockwell, Jr. J.W., 2002, CWP/SU: Seismic Unix Release No. 36: a free
package for seismic research and processing, Center for Wave Phenomena,
Colorado School of Mines.
Lamb, L., 1990, Learning the vi Editor, Sebastopol, California: O’Reilly & Associates,
Inc..
Lindseth, R.O., 1982, Digital Processing of Geophysical Data – A Review, Tulsa: Society
of Exploration Geophysicists (reprint).
Moore, G.F., Shipley, T.H., Stoffa, P.L., Karig, D.E., Taira, A., Kuramoto, S.,
Tokuyama, H., Suyehiro, K., 1990, Journal of Geopysical Research, Vol. 95, No.
B6, 8753-8765.
Robinson, E.A., Treitel, S., 2000, Geophysical Signal Analysis (reprint), Tulsa: Society
of Exploration Geophysicists, p. 278-279.
Sheriff, R.E., 2002, Encyclopedic Dictionary of Applied Geophysics, 4th Edition, Tulsa:
Society of Exploration Geophysicists.
Stockwell, Jr. J.W., & Cohen, J.K., 2002, The New SU User’s Manual, Golden: CWP,
Colorado School of Mines.
Yilmaz, Ö., 1987, Seismic Data Processing, Tulsa: Society of Exploration Geophysicists.
Yilmaz, Ö., 2001, Seismic Data Analysis, Tulsa: Society of Exploration Geophysicists.

R-1
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

This page has been intentionally left blank

 Appendix A: Seismic Un*x Legal Statement

Appendix A: Seismic Un*x Legal Statement

Below, presented verbatim, is the Legal Statement that accompanies Seismic Un*x
release 38.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

This file is property of the Colorado School of Mines.

Copyright © 2004, Colorado School of Mines, all rights reserved.

Warranty Disclaimer:
NO GUARANTEES, OR WARRANTIES, EITHER EXPRESS OR IMPLIED, ARE
PROVIDED BY CWP, CSM, ANY EMPLOYEE OR MEMBER OF THE AFORESAID
ORGANIZATIONS, OR BY ANY CONTRIBUTOR TO THIS SOFTWARE
PACKAGE, REGARDING THE ACCURACY, SAFETY, FITNESS FOR A
PARTICULAR PURPOSE, OR ANY OTHER QUALITY OR CHARACTERISTIC OF
THIS SOFTWARE, OR ANY DERIVATIVE SOFTWARE.

Export Restriction Disclaimer:

We believe that CWP/SU: Seismic Un*x is a low technology product that does not
appear on the Department of Commerce CCL list of restricted exports. Accordingly, we
believe that our product meets the qualifications of an ECCN (export control
classification number) of EAR99 and we believe it fits the qualifications of NRR (no
restrictions required), and is thus not subject to export restrictions of any variety.

Limited License:
The CWP/SU Seismic Un*x package (SU) is not public domain software, but it is
available free under the following terms and conditions:

1. Permission to use, copy, and modify this software for any purpose without fee and
within the guidelines set forth below is hereby granted, provided that the above copyright
notice, the warranty disclaimer, and this permission notice appear in all copies, and the
name of the Colorado School of Mines (CSM) not be used in advertising or publicity
pertaining to this software without the specific, written permission of CSM.

2. The simple repackaging and selling of the SU package as is, as a commercial software
product, is expressly forbidden without the prior written permission of CSM. Any
approved repackaging arrangement will carry the following restriction: only a modest
profit over reproduction costs may be realized by the reproducer.

Approved Reference Format:

In publications, please refer to SU as per the following example:

A-1
 Appendix A: Seismic Un*x Legal Statement

Cohen, J. K. and Stockwell, Jr. J. W., (200_), CWP/SU: Seismic Un*x Release No. __: a
free package for seismic research and processing, Center for Wave Phenomena, Colorado
School of Mines.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Recent articles about SU in peer-reviewed journals:

• Saeki, T., (1999), A guide to Seismic Un*x (SU)(2)---examples of data processing
(part 1), data input and preparation of headers, Butsuri-Tansa (Geophysical
Exploration), vol. 52, no. 5, 465-477.
• Stockwell, Jr. J. W. (1999), The CWP/SU: Seismic Un*x Package, Computers and
Geosciences, May 1999.
• Stockwell, Jr. J. W. (1997), Free Software in Education: A case study of CWP/SU:
Seismic Un*x, The Leading Edge, July 1997.
• Templeton, M. E., Gough, C.A., (1998), Web Seismic Un*x: Making seismic
reflection processing more accessible, Computers and Geosciences, in press.

Acknowledgements:
SU stands for CWP/SU:Seismic Un*x, a processing line developed at Colorado School of
Mines, partially based on Stanford Exploration Project (SEP) software.

A-2
 Appendix B: Seismic Un*x at MTU

Appendix B: Seismic Un*x at Michigan Tech

The use of SU on any given system depends on the “path” that is assigned by the
system administrator. This Appendix is an example for system administrators who may
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

want to emulate the approach we use at Michigan Technological University.

B.1 Using Seismic Un*x

At Michigan Tech, you must use SU within a special terminal (x-term) window. Do
the following:
1. Log into a Unix machine.
2. Open an x-term window and enter seismicx. This opens another window, titled
Seismic Un*x #, where # is your SU version number.
The first time you do this, it will take a while, because you are creating a new
subdirectory structure.
3. In the Seismic Un*x window, enter pwd (on some computers you have to use
cwd) to see the directory in which you are now working; it should be in your
home area, in a directory called seismicx.
4. Enter ls to see a listing of the files and subdirectories in that directory. You
should see a subdirectory called demos.
5. Enter cd demos to change to that subdirectory, then enter ls to see a listing
within it. When you want to move back up one level of directory,
enter cd .. (Yes, that is: cd space dot dot. Note that if you enter cd
by itself, you will go to your home directory.).
After you have worked in the demos directory for a while, if you decide you have
made a mess of the demos directory, remove the directory and all its contents, then close
the Seismic Un*x window. The next time you open a Seismic Un*x window, the system
will create a fresh demos directory.

B.2 Terminal Windows

You must open a Seismic Un*x x-term window and work within it to execute SU
commands. If you are not in a Seismic Un*x window, your SU commands will not work,
even if you are in the seismicx directory (or any of its subdirectories). This is because the
setup procedure that places the SU package in the user’s “path” is only in place when the
user is in an x-term window also created by the setup procedure.

B-1
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

This page has been intentionally left blank

 Appendix C: Utility Programs

C. Utility Programs
C.1 Introduction
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Sometimes the library of Seismic Unix (SU) programs doesn’t have exactly what we
want. Sometimes it is just fun to write a program that does exactly what we want. This
Appendix presents three C programs that do simple tasks in SU.
• sukeycount.c
• sudumptrace.c
• tvnmoqc.c
These three programs were written while we used Seismic Un*x version 38. You
might find these programs in a later SU version. We are glad to contribute to the SU
library.
Before we examine these programs, we recommend the programming sections of
“The New SU User’s Manual” (Stockwell & Cohen, 2002). The Manual web site is:
http://www.cwp.mines.edu/sututor/sututor.html
The Manual’s Contents web page is:
http://www.cwp.mines.edu/sututor/node1.html
The Manual web page that describes shell programming, “Extending SU by shell
programming,” is:
http://www.cwp.mines.edu/sututor/node128.html
This Appendix does not discuss shell programming. We think you have seen many
shell examples in previous chapters. However, we recommend the above link because it
helped us learn SU shell programming.
The SU web page that begins the section about C programming, “How to Write an
SU Program,” is:
http://www.cwp.mines.edu/sututor/node136.html
Under this section, there are three web pages:
• “Setting up the Makefile” –
http://www.cwp.mines.edu/sututor/node137.html
• “A template SU program” –
http://www.cwp.mines.edu/sututor/node138.html
• “Writing a new program: suvlength” –
http://www.cwp.mines.edu/sututor/node139.html
You will understand the examples in this Appendix faster if you first study the
Manual pages. Reading and understanding the Makefile web page will save you the effort
of compiling and linking your program – the Makefile helps you compile and link.
In this Appendix, we do not teach C programming; we presume you are familiar with
C. This Appendix is written to help you extend your knowledge of C into SU. However,
we do not explain our programs in detail. We hope that our simple examples help you
start on the road to becoming a C programmer in SU.

C-1
 Appendix C: Utility Programs

C.2 Program sukeycount

Below is the selfdoc of sukeycount: The user parameter of sukeycount is a key.
SUKEYCOUNT - sukeycount writes a count of a selected key
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

sukeycount key=key1 < infile [> outfile]

Required parameters:
key=key1 One key word.

Optional parameters:
verbose=0 quiet
=1 chatty

Writes the key and the count to the terminal or a text

file when a change of key occurs. This does not provide
a unique key count (see SUCOUNTKEY for that).
Note that for key values 1 2 3 4 2 5
value 2 is counted once per occurrence since this program
only recognizes a change of key, not total occurrence.

Examples:
sukeycount < stdin key=fldr
sukeycount < stdin key=fldr > out.txt

Program sukeycount reads the first trace. After that, each trace’s key value is
compared against the previous trace. If the new trace’s key value is the same as the
previous key value, a counter is incremented. The next trace is read and the comparison is
made again. When the key value changes, sukeycount writes (to the screen or to a log
file) the key name, the value that was recently tested, and the count of the traces with that
value. After each write, the counter is re-set.
If the input data are a 2-D line sorted in CMP-order, a count of traces within CMPs is
equivalent to a fold count. Suppose the first 16 traces of cmp4.su have the following cdp
values:
sequential
cdp
trace
value
number
1 1
2 2
3 3
4 3
5 4
6 4
7 5
8 5
9 5
10 6
11 6
12 6
13 7
14 7
15 7
16 7
Below, we use sukeycount on the CMP-ordered 2-D data.
sukeycount < cmp4.su key=cdp > model4Fold.txt
Below are the first 7 lines of output

C-2
 Appendix C: Utility Programs

cdp = 1 has 1 trace(s)

cdp = 2 has 1 trace(s)
cdp = 3 has 2 trace(s)
cdp = 4 has 2 trace(s)
cdp = 5 has 3 trace(s)
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

cdp = 6 has 3 trace(s)

cdp = 7 has 4 trace(s)
To generate Figure 7.2, we
• opened Excel,
• imported file model4Fold.txt,
• specified that the imported file is space-delimited, and
• used Excel’s Chart Wizard to plot the two numeric columns.
Below is program sukeycount.c.
1 /* David Forel, Jan. 2005.*/
2
3 #include "su.h"
4 #include "segy.h"
5 #include "header.h"
6
7 /*********************** self documentation **********************/
8 char *sdoc[] = {
9 " ",
10 " SUKEYCOUNT - sukeycount writes a count of a selected key ",
11 " ",
12 " sukeycount key=key1 < infile [> outfile] ",
13 " ",
14 " Required parameters: ",
15 " key=key1 One key word. ",
16 " ",
17 " Optional parameters: ",
18 " verbose=0 quiet ",
19 " =1 chatty ",
20 " ",
21 " Writes the key and the count to the terminal or a text ",
22 " file when a change of key occurs. This does not provide ",
23 " a unique key count (see SUCOUNTKEY for that). ",
24 " Note that for key values 1 2 3 4 2 5 ",
25 " value 2 is counted once per occurrence since this program ",
26 " only recognizes a change of key, not total occurrence. ",
27 " ",
28 " Examples: ",
29 " sukeycount < stdin key=fldr ",
30 " sukeycount < stdin key=fldr > out.txt ",
31 " ",
32 NULL};
33
34 /* Credits:
35 *
36 * MTU: David Forel
37 */
38 /**************** end self doc ***********************************/
39
40 segy tr ;
41
42 int
43 main(int argc, char **argv)
44 {
45 cwp_String key[SU_NKEYS] ; /* array of keywords */
46 int nkeys ; /* number of keywords to retrieve */

C-3
 Appendix C: Utility Programs

47 int iarg ; /* arguments in argv loop */

48 int countkey = 0 ; /* counter of keywords in argc loop */
49 cwp_String output ; /* string representing output format */
50 int verbose = 0 ; /* verbose ? */
51 cwp_String type1 ; /* key string */
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

52 float sort, sortold ; /* for comparing new/old key values */

53 int isort, isortold ; /* for comparing new/old key values */
54 int gatherkount ; /* counter of traces within key */
55 int itotal = 0 ; /* counter of total traces */
56
57 /* Initialize */
58 initargs(argc, argv) ;
59 requestdoc(1) ;
60
61 sortold = -99999. ;
62 isortold = -99999 ;
63
64 /* Get key values */
65 if (!getparint("verbose",&verbose)) verbose=0 ;
66 if ((nkeys=countparval("key"))!=0)
67 {
68 getparstringarray("key",key) ;
69 }
70 else
71 {
72 /* support old fashioned method for inputting key fields */
73 /* as single arguments: sukeycount key1 */
74 if (argc==1) err("must set one key value!") ;
75
76 for (iarg = 1; iarg < argc; ++iarg)
77 {
78 cwp_String keyword ; /* keyword */
79
80 keyword = argv[iarg] ;
81
82 if (verbose) warn("argv=%s",argv[iarg]);
83 /* get array of types and indexes to be set */
84 if ((strncmp(keyword,"output=",7)!=0))
85 {
86 key[countkey] = keyword ;
87 ++countkey ;
88 }
89 if (countkey==0) err("must set one key value!") ;
90 if (countkey>1) err("must set only one key value!") ;
91 }
92 nkeys=countkey;
93 }
94 if (nkeys>1) err("must set only one key value!") ;
95
96 printf("\n") ;
97
98 /* Loop over traces */
99 gatherkount = 0 ;
100 while (gettr(&tr))
101 {
102 /* Do not loop over keys because only one is used */
103 Value vsort ;
104 gethdval(&tr, key[0], &vsort) ;
105 type1 = hdtype(key[0]) ;
106
107 if (*type1 == 'f') /* float header */
108 {
109 sort = vtof(type1,vsort) ;

C-4
 Appendix C: Utility Programs

110
111 /* Don't write just because first trace is new */
112 if ( itotal == 0 ) sortold = sort ;
113
114 if ( sort != sortold )
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

115 {
116 printf(" %8s = %f", key[0], sortold) ;
117 printf(" has %d trace(s)", gatherkount) ;
118 printf("\n") ;
119 sortold = sort ;
120 gatherkount = 0 ;
121 }
122 ++gatherkount ;
123 ++itotal ;
124 }
125 else /* non-float header */
126 {
127 isort = vtoi(type1,vsort) ;
128
129 /* Don't write just because first trace is new */
130 if ( itotal == 0 ) isortold = isort ;
131
132 if ( isort != isortold )
133 {
134 printf(" %8s = %d", key[0], isortold) ;
135 printf(" has %d trace(s)", gatherkount) ;
136 printf("\n") ;
137 isortold = isort ;
138 gatherkount = 0 ;
139 }
140 ++gatherkount ;
141 ++itotal ;
142 }
143 }
144
145 /* Write after last trace is read */
146 if (*type1 == 'f')
147 {
148 printf(" %8s = %f", key[0], sortold) ;
149 printf(" has %d trace(s)", gatherkount) ;
150 printf("\n") ;
151 }
152 else
153 {
154 printf(" %8s = %d", key[0], isortold) ;
155 printf(" has %d trace(s)", gatherkount) ;
156 printf("\n") ;
157 }
158 printf("\n %d trace(s) read\n\n", itotal) ;
159
160 return(CWP_Exit()) ;
161 }
162

C.3 Program sudumptrace

A Seismic Un*x program to print trace amplitude values is suascii. (It also prints
trace headers.) For each trace value, suascii prints two columns: the first column is a time
count, the second column is the amplitude value. However, if the amplitude values for
each trace were printed in separate columns, we could more easily compare the traces.

C-5
 Appendix C: Utility Programs

Below is the selfdoc of sudumptrace: The default value num=4 means sudumptrace
prints values for the first four traces it reads. No key names are required. If the user
supplies key names, key names and their values are printed above each trace. If hpf=1,
the print format of trace key values is exponential; the default format is floating point.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Trace amplitudes are printed using exponential format.

SUDUMPTRACE - print selected header values and data.
Print first num traces.
Use SUWIND to skip traces.

sudumptrace < stdin [> ascii_file]

Optional parameters:
num=4 number of traces to dump
key=key1,key2,... key(s) to print above trace values
hpf=0 header print format is float
=1 print format is exponential

Examples:
sudumptrace < inseis.su PRINTS: 4 traces, no headers
sudumptrace < inseis.su key=tracf,offset
sudumptrace < inseis.su num=7 key=tracf,offset > info.txt
sudumptrace < inseis.su num=7 key=tracf,offset hpf=1 > info.txt

In Section 12.3, we used sudumptrace to examine trace amplitudes of Taiwan shot

gather (fldr) 930. We also used suwind to insure that the first traces into sudumptrace
were the ones we wanted to examine.
Here, we use a small data set mini3.su. Below is the surange output of mini3.su.
surange < mini3.su
7 traces:
tracl=(129169,129175) tracr=(14401,14407) fldr=900 tracf=(1,7) ep=840
cdp=(4419,4425) cdpt=(1,7) trid=1 nhs=1 offset=(-3663,-3513)
sdepth=80000 swdep=3110000 scalel=-10000 scalco=1 sx=-4450200
gx=(-4110870,-4096381) gy=(90923,94805) counit=3 tstat=12 delrt=28
ns=19 dt=4000 gain=9 afilf=160 afils=72
lcf=3 hcf=160 lcs=6 hcs=72 year=95
day=260 hour=12 minute=58 sec=40
File mini3.su has 7 traces and 19 time samples (ns=19). Also note that the delay time is
28 ms (delrt=28). Below, we use sudumptrace on mini3.su.
sudumptrace < mini3.su key=tracl,cdp,tracf,ns

num traces = 4 num samples = 19

tracl 129169.0000 129170.0000 129171.0000 129172.0000

cdp 4419.0000 4420.0000 4421.0000 4422.0000
tracf 1.0000 2.0000 3.0000 4.0000
ns 19.0000 19.0000 19.0000 19.0000

Counter Time Values

1 0.032 -4.6336e-01 1.5644e-01 -1.2573e-01 5.4677e-02
2 0.036 -1.9764e+00 6.2943e-01 -5.3233e-01 2.4022e-01
3 0.040 -4.0162e+00 1.1478e+00 -1.0636e+00 4.8795e-01
4 0.044 -4.1991e+00 9.5269e-01 -1.0558e+00 4.4756e-01
5 0.048 -3.1962e+00 5.0495e-01 -7.0221e-01 2.0370e-01
6 0.052 -3.4293e+00 4.8322e-01 -6.9947e-01 2.3813e-01
7 0.056 -3.3138e+00 2.8859e-01 -6.4975e-01 3.6629e-01
8 0.060 -2.5544e+00 5.0591e-01 -4.7939e-01 3.4599e-01

C-6
 Appendix C: Utility Programs

9 0.064 -2.7929e+00 1.6010e+00 -5.5672e-01 4.0045e-01

10 0.068 -3.0077e+00 2.1449e+00 -5.4405e-01 4.9452e-01
11 0.072 -2.7386e+00 1.6686e+00 -5.5408e-01 5.8649e-01
12 0.076 -2.0205e+00 1.0067e+00 -7.8985e-01 7.6442e-01
13 0.080 -1.1539e+00 1.0423e+00 -9.8582e-01 9.8357e-01
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

14 0.084 -1.0390e+00 1.2934e+00 -1.0409e+00 1.1426e+00

15 0.088 -9.2735e-01 8.8366e-01 -8.3712e-01 1.2262e+00
16 0.092 -5.6792e-01 5.5448e-01 -5.3082e-01 1.3517e+00
17 0.096 -5.5705e-01 5.7637e-01 -5.0095e-01 1.4079e+00
18 0.100 -6.4979e-01 9.0827e-01 -5.6816e-01 1.4072e+00
19 0.104 -5.0397e-01 1.6227e+00 -3.2804e-01 1.5339e+00
The first output column is time sample count; the second column is time value
(seconds). The next four columns are amplitude values of the first four input traces. The
amplitude columns are unevenly spaced because they are tab-delimited.
The plot of mini3.su, Figure C.1, uses key=tracf to label the x-axis.

Figure C.1: Plot of mini3.su.

For another test, we window mini3.su so the input to sudumptrace is the last 3 traces
of the file.
suwind key=tracf min=5 < mini3.su | sudumptrace key=tracf,cdp,offset
Although we used the default num=4, only the last three traces were available for
input to sudumptrace. Below is the screen output.

C-7
 Appendix C: Utility Programs

num traces = 3 num samples = 19

tracf 5.0000 6.0000 7.0000

cdp 4423.0000 4424.0000 4425.0000
offset -3563.0000 -3538.0000 -3513.0000
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Counter Time Values

1 0.032 -9.5147e-02 9.3358e-01 -2.3633e-01
2 0.036 -3.1833e-01 3.8977e+00 -9.3778e-01
3 0.040 -4.1830e-01 7.5830e+00 -1.5560e+00
4 0.044 -2.4751e-01 7.1671e+00 -6.5207e-01
5 0.048 -4.9481e-01 4.6742e+00 7.9582e-01
6 0.052 -8.2958e-01 5.2009e+00 8.2085e-01
7 0.056 -3.3455e-01 5.1407e+00 8.2723e-01
8 0.060 -1.7911e-01 4.1310e+00 1.2068e+00
9 0.064 -5.3361e-01 4.7825e+00 3.2759e-01
10 0.068 -8.2213e-02 4.2018e+00 -7.1798e-01
11 0.072 4.3474e-01 3.5753e+00 -9.7540e-01
12 0.076 2.3809e-01 4.1400e+00 -1.1255e+00
13 0.080 1.3812e-02 3.8605e+00 -9.5250e-01
14 0.084 1.2431e-01 3.4150e+00 -5.5377e-01
15 0.088 1.6596e-01 3.1037e+00 -5.0291e-01
16 0.092 4.8231e-03 2.4580e+00 -3.4983e-01
17 0.096 -2.8939e-02 2.1375e+00 2.9766e-01
18 0.100 2.4795e-01 2.7438e+00 1.0379e+00
19 0.104 7.8749e-01 2.6644e+00 1.3592e+00

Below is program sudumptrace.c.

1 /* David Forel Jan 2005 */
2
3 #include "su.h"
4 #include "segy.h"
5 #include "header.h"
6 #include <signal.h>
7
8 /*********************** self documentation **********************/
9 char *sdoc[] = {
10 " ",
11 " SUDUMPTRACE - print selected header values and data. ",
12 " Print first num traces. ",
13 " Use SUWIND to skip traces. ",
14 " ",
15 " sudumptrace < stdin [> ascii_file] ",
16 " ",
17 " Optional parameters: ",
18 " num=4 number of traces to dump ",
19 " key=key1,key2,... key(s) to print above trace values ",
20 " hpf=0 header print format is float ",
21 " =1 print format is exponential ",
22 " ",
23 " Examples: ",
24 " sudumptrace < inseis.su PRINTS: 4 traces, no headers ",
25 " sudumptrace < inseis.su key=tracf,offset ",
26 " sudumptrace < inseis.su num=7 key=tracf,offset > info.txt ",
27 " sudumptrace < inseis.su num=7 key=tracf,offset hpf=1 > info.txt ",
28 " ",
29 NULL};
30
31 /* Credits:
32 * MTU: David Forel
33 *

C-8
 Appendix C: Utility Programs

34 * Trace header field accessed: nt, dt, delrt

35 */
36 /**************** end self doc ***********************************/
37
38 /* subroutine prototypes */
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

39 void dump(float *data, float dt, float *hedr, cwp_String *key,

40 float delrt, int nkeys, int ntr, int nt, int hpf) ;
41 static void closefiles(void) ;
42
43 /* Globals (so can trap signal) defining temporary disk files */
44 char tracefile[BUFSIZ] ; /* filename for the file of traces */
45 char headerfile[BUFSIZ] ; /* filename for the file of headers */
46 FILE *tracefp ; /* fp for trace storage file */
47 FILE *headerfp ; /* fp for header storage file */
48
49 segy tr ;
50
51 int
52 main(int argc, char **argv)
53 {
54 char *tmpdir ; /* directory path for tmp files */
55 cwp_Bool istmpdir=cwp_false ; /* true for user given path */
56 float *hedr ; /* the headers */
57 float *data ; /* the data */
58
59 int nt ; /* number of trace samples */
60 float dt ; /* sample interval, sec */
61 float delrt ; /* delay recording time, sec */
62 cwp_String key[SU_NKEYS] ; /* array of keywords */
63 cwp_String type ; /* key string type */
64 int nkeys ; /* number of keywords */
65 int ikey, itr, ntr = 0 ; /* counters */
66 int num ; /* number of traces to dump */
67 int numtr = 4 ; /* number of traces to dump */
68 int hpf ; /* header print format */
69
70 /* Initialize */
71 initargs(argc, argv) ;
72 requestdoc(1) ;
73
74 /* Look for user-supplied tmpdir */
75 if (!getparstring("tmpdir",&tmpdir) &&
76 !(tmpdir = getenv("CWP_TMPDIR"))) tmpdir="";
77 if (!STREQ(tmpdir, "") && access(tmpdir, WRITE_OK))
78 err("you can't write in %s (or it doesn't exist)", tmpdir);
79
80 /* Get values from first trace */
81 if (!gettr(&tr)) err("can't get first trace");
82 nt = (int) tr.ns ; /* Get nt */
83 dt = ((double) tr.dt)/1000000.0 ; /* microsecs to secs */
84 if (!dt) getparfloat("dt", &dt) ;
85 if (!dt) MUSTGETPARFLOAT("dt", &dt) ;
86 delrt = ((double) tr.delrt)/1000.0 ; /* millisecs to secs */
87
88 /* Get parameters */
89 if (getparint ("num", &num)) numtr = num ;
90 if ((nkeys=countparval("key"))!=0) getparstringarray("key",key) ;
91 hedr = ealloc1float(nkeys*numtr) ; /* make space for headers */
92 if (!getparint ("hpf", &hpf)) hpf = 0 ;
93
94 /* Store traces, headers in tempfiles */
95 if (STREQ(tmpdir,""))
96 {

C-9
 Appendix C: Utility Programs

97 tracefp = etmpfile();
98 headerfp = etmpfile();
99
100 do
101 {
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

102 ++ntr;
103 efwrite(&tr, HDRBYTES, 1, headerfp);
104 efwrite(tr.data, FSIZE, nt, tracefp);
105
106 /* Get header values */
107 for (ikey=0; ikey<nkeys; ++ikey)
108 {
109 Value val;
110 float fval;
111
112 gethdval(&tr, key[ikey], &val) ;
113 type = hdtype(key[ikey]) ;
114 fval = vtof(type,val) ;
115 hedr[(ntr-1)*nkeys+ikey] = fval ;
116 }
117
118 }
119 while (ntr<numtr && gettr(&tr)) ;
120
121 }
122 else /* user-supplied tmpdir */
123 {
124 char directory[BUFSIZ];
125 strcpy(directory, tmpdir);
126 strcpy(tracefile, temporary_filename(directory));
127 strcpy(headerfile, temporary_filename(directory));
128 /* Handle user interrupts */
129 signal(SIGINT, (void (*) (int)) closefiles);
130 signal(SIGQUIT, (void (*) (int)) closefiles);
131 signal(SIGHUP, (void (*) (int)) closefiles);
132 signal(SIGTERM, (void (*) (int)) closefiles);
133 tracefp = efopen(tracefile, "w+");
134 headerfp = efopen(headerfile, "w+");
135 istmpdir=cwp_true;
136
137 do
138 {
139 ++ntr;
140 efwrite(&tr, HDRBYTES, 1, headerfp);
141 efwrite(tr.data, FSIZE, nt, tracefp);
142
143 /* Get header values */
144 for (ikey=0; ikey<nkeys; ++ikey)
145 {
146 Value val;
147 float fval;
148
149 gethdval(&tr, key[ikey], &val) ;
150 type = hdtype(key[ikey]) ;
151 fval = vtof(type,val) ;
152 hedr[(ntr-1)*nkeys+ikey] = fval ;
153 }
154
155 }
156 while (ntr<numtr && gettr(&tr)) ;
157
158 }
159

C-10
 Appendix C: Utility Programs

160 /* Rewind after read, allocate space */

161 erewind(tracefp);
162 erewind(headerfp);
163 data = ealloc1float(nt*ntr);
164
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

165 /* Load traces into data and close tmpfile */

166 efread(data, FSIZE, nt*ntr, tracefp);
167 efclose(tracefp);
168 if (istmpdir) eremove(tracefile);
169
170 rewind(headerfp);
171 rewind(tracefp);
172
173 /* Do trace work */
174 dump(data, dt, hedr, key, delrt, nkeys, ntr, nt, hpf) ;
175
176 /* close */
177 efclose(headerfp);
178 if (istmpdir) eremove(headerfile);
179
180 free1(hedr) ;
181 free1(data) ;
182
183 return(CWP_Exit()) ;
184 }
185
186
187 void dump(float *data, float dt, float *hedr, cwp_String *key,
188 float delrt, int nkeys, int ntr, int ntime, int hpf)
189 /*
190 Dump headers and traces in column format; one trace per column.
191 INPUT:
192 data array of trace data values
193 dt trace sample interval
194 delrt delay recording time, sec
195 hedr array of trace headers
196 key array of key names
197 nkeys number of keys (headers) to print above trace values
198 ntr number of traces to dump
199 ntime number of time samples on trace
200 hpf header print format flag: 0=float, 1=exponential
201 OUTPUT: none
202 */
203 {
204 int i, ikey, j, k, m ; /* counters */
205
206 printf("\nnum traces = %d num samples = %d \n\n",ntr,ntime) ;
207
208 for (k=0; k<nkeys; ++k) /* Print headers */
209 {
210 printf("%17s ", key[k]) ;
211 for (m=0; m<ntr; ++m)
212 {
213 if (hpf==0)
214 {
215 printf("%11.4f\t", hedr[m*nkeys + k]) ;
216 }
217 else
218 {
219 printf("%11.4e\t", hedr[m*nkeys + k]) ;
220 }
221 }
222 putchar('\n') ;

C-11
 Appendix C: Utility Programs

223 }
224
225 putchar('\n') ;
226
227 printf("\nCounter Time Values\n") ; /* Column titles */
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

228
229 for (i=1; i<=ntime; ++i) /* Print trace values */
230 {
231 printf(" %6d ", i) ;
232 printf(" %8.3f ", dt*(i)+delrt) ;
233 for (j=1; j<=ntr; ++j)
234 {
235 printf("%11.4e\t", data[(j-1)*ntime+(i-1)]) ;
236 }
237 putchar('\n') ;
238 }
239 putchar('\n') ;
240
241 }
242
243
244 /* for graceful interrupt termination */
245 static void closefiles(void)
246 {
247 efclose(headerfp);
248 efclose(tracefp);
249 eremove(headerfile);
250 eremove(tracefile);
251 exit(EXIT_FAILURE);
252 }
253

C.4 Program tvnmoqc

Program suximage can put lines on the image when the x-y line values are available
as two columns. The input to sunmo, a file of CDP values and associated row pairs of
tnmo-vnmo values (see Section 8.2) can provide values for an suximage line overlay. The
task is to overlay a CDP’s time-velocity values on the suximage velan. The difficulty is
that the sunmo rows must be converted to columns for suximage.
Using mode=2, tvnmoqc converts a CDP’s corresponding tnmo-vnmo rows to
columns. The output file is named by a prefix that is supplied by the user (required
parameter prefix when mode=2); the output file name suffix is the CDP number. For
example, in line 15 of the code (below), if the user supplies prefix=velqc, the first output
file is named velqc.15 and the second output file is named velqc.35.
In either mode 1 or 2, tvnmoqc also makes several checks that sunmo does – most of
the parameter checks were taken from sunmo.
• CDP values must be supplied.
• There must be the same number of tnmo and vnmo series as there are CDPs.
• And several other checks that sunmo does.
One important difference between tvnmoqc and sunmo is that tvnmoqc does not
stop if a time series does not always increase. If a time series pair is found that does not
increase, a message is printed to the screen and processing continues. You can use
tvnmoqc as a quality check of the cdp-tnmo-vnmo values before using sunmo.
Section 8.2 shows examples of using mode=1 and mode=2.

C-12
 Appendix C: Utility Programs

In tvnmoqc.c:
Lines 99-139 were taken almost line-by-line from sunmo. These lines check that the
cdp-tnmo-vnmo values have acceptable values for sunmo.
Lines 137-138 are the diagnostic print to the screen.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Line 143 creates the output file name.

Line 144 calls function mktvfile that (1) converts the tnmo-vnmo values from rows to
columns and (2) writes the t-v columns to the file.
1 /* David Forel May 2005 */
2
3 #include "par.h"
4 #include <string.h>
5
6 /*********************** self documentation **************************/
7 char *sdoc[] = {
8 " ",
9 " TVNMOQC - Check tnmo-vnmo pairs; create t-v column files ",
10 " ",
11 " tvnmoqc [parameters] cdp=... tnmo=... vnmo=... ",
12 " ",
13 " Example: ",
14 " tvnmoqc mode=1 \\ ",
15 " cdp=15,35 \\ ",
16 " tnmo=0.0091,0.2501,0.5001,0.7501,0.9941 \\ ",
17 " vnmo=1497.0,2000.0,2500.0,3000.0,3500.0 \\ ",
18 " tnmo=0.0082,0.2402,0.4902,0.7402,0.9842 \\ ",
19 " vnmo=1495.0,1900.0,2400.0,2900.0,3400.0 ",
20 " ",
21 " Required Parameter: ",
22 " prefix= Prefix of output t-v file(s) ",
23 " Required only for mode=2 ",
24 " ",
25 " Optional Parameter: ",
26 " mode=1 1=qc: check that tnmo values increase ",
27 " 2=qc and output t-v files ",
28 " ",
29 " mode=1 ",
30 " TVNMOQC checks that there is a tnmo and vnmo series for each CDP ",
31 " and it checks that each tnmo series increases in time. ",
32 " ",
33 " mode=2 ",
34 " TVNMOQC does mode=1 checking, plus ... ",
35 " ",
36 " TVNMOQC converts par (MKPARFILE) values written as: ",
37 " ",
38 " cdp=15,35,...,95 \\ ",
39 " tnmo=t151,t152,...,t15n \\ ",
40 " vnmo=v151,v152,...,v15n \\ ",
41 " tnmo=t351,t352,...,t35n \\ ",
42 " vnmo=v351,v352,...,v35n \\ ",
43 " tnmo=... \\ ",
44 " vnmo=... \\ ",
45 " tnmo=t951,t952,...,t95n \\ ",
46 " vnmo=v951,v952,...,v95n \\ ",
47 " ",
48 " to column format. The format of each output file is: ",
49 " ",
50 " t1 v1 ",
51 " t2 v2 ",
52 " ... ",

C-13
 Appendix C: Utility Programs

53 " tn vn ",
54 " ",
55 " One file is output for each input pair of tnmo-vnmo series. ",
56 " ",
57 " A CDP VALUE MUST BE SUPPLIED FOR EACH TNMO-VNMO ROW PAIR. ",
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

58 " ",
59 " Prefix of each output file is the user-supplied value of ",
60 " parameter PREFIX. ",
61 " Suffix of each output file is the cdp value. ",
62 " For the example above, output files names are: ",
63 " PREFIX.15 PREFIX.35 ... PREFIX.95 ",
64 " ",
65 NULL};
66
67 /* Credits:
68 * MTU: David Forel (adapted from SUNMO)
69 */
70 /**************** end self doc ***************************************/
71
72 void mktvfile(char outfile[], int ntnmo, float *tnmo, float *vnmo);
73 int main(int argc, char **argv)
74 {
75 int k; /* index used in loop */
76 int mode; /* mode=1: qc; mode=2: qc + make cdp-t-v file */
77 int icdp; /* index into cdp array */
78 int ncdp; /* number of cdps specified */
79 int *cdp; /* array[ncdp] of cdps */
80 int nvnmo; /* number of vnmos specified */
81 float *vnmo; /* array[nvnmo] of vnmos */
82 int ntnmo; /* number of tnmos specified */
83 float *tnmo; /* array[ntnmo] of tnmos */
84 cwp_String prefix; /* prefix of output files */
85 char dot[] = "."; /* for output file name */
86 char outfile[80]; /* output file name */
87
88 /* Hook up getpar */
89 initargs(argc, argv);
90 requestdoc(1);
91
92 /* Get parameters */
93 if(!getparint("mode",&mode))mode=1;
94 if(mode==2)
95 if (!getparstring("prefix", &prefix))
96 err("When mode=2, you must supply a prefix name.");
97
98 /* Are there cdp values and vnmo-tnmo sets for each cdp? */
99 ncdp = countparval("cdp");
100 warn("This file has %i CDPs.",ncdp);
101 if (ncdp>0) {
102 if (countparname("vnmo")!=ncdp)
103 err("A vnmo set must be specified for each cdp");
104 if (countparname("tnmo")!=ncdp)
105 err("A tnmo set must be specified for each cdp");
106 } else {
107 err("A cdp value must be supplied for each tnmo-vnmo set");
108 }
109
110 /* Get cdp values */
111 cdp = ealloc1int(ncdp);
112 if (!getparint("cdp",cdp))
113 err("A cdp value must be supplied for each tnmo-vnmo set");
114
115 /* Get tnmo-vnmo values */

C-14
 Appendix C: Utility Programs

116 for (icdp=0; icdp<ncdp; ++icdp)

117 {
118 nvnmo = countnparval(icdp+1,"vnmo");
119 ntnmo = countnparval(icdp+1,"tnmo");
120 if (nvnmo!=ntnmo)
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

121 err("number of vnmo and tnmo values must be equal");

122 if (nvnmo==0) err("Each cdp must have at least one velocity");
123 if (ntnmo==0) err("Each cdp must have at least one time");
124
125 vnmo = ealloc1float(nvnmo);
126 tnmo = ealloc1float(nvnmo);
127
128 if (!getnparfloat(icdp+1,"vnmo",vnmo))
129 err("Each cdp must have at least one velocity");
130 if (!getnparfloat(icdp+1,"tnmo",tnmo))
131 err("Each cdp must have at least one time");
132 for (k=1; k<ntnmo; ++k)
133 if (tnmo[k]<=tnmo[k-1]) {
134 warn("tnmo values must increase for use in NMO");
135 /* A tnmo series that does not increase is not an
136 error in velocity QC (but IS an error in SUNMO) */
137 warn("For cdp=%i, check times %g and %g.",
138 cdp[icdp],tnmo[k-1],tnmo[k]);
139 }
140
141 /* write cdp-tnmo-vnmo values to file "prefix" "dot" "cdp" */
142 if (mode == 2) {
143 sprintf(outfile,"%s%s%i",prefix,dot,cdp[icdp]);
144 mktvfile(outfile,ntnmo,tnmo,vnmo);
145 }
146
147 free1float(vnmo);
148 free1float(tnmo);
149 }
150
151 free1int(cdp);
152 warn("End of cdp-tnmo-vnmo check.");
153
154 return(CWP_Exit());
155 }
156
157
158 void mktvfile(char outfile[], int ntnmo, float *tnmo, float *vnmo)
159 {
160 int i; /* index used in loop */
161 FILE *fptr; /* file pointer for output */
162
163 fptr = efopen(outfile, "w");
164 for (i=0; i<ntnmo; ++i) {
165 fprintf(fptr,"%g %g\n",tnmo[i],vnmo[i]);
166 }
167 efclose(fptr);
168 }
169

C.5 Conclusion
As the SU User’s Manual states, “The secret to efficient SU coding is finding an
existing program similar to the one you want to write.” For example, if you want part of
your program to modify trace headers (keys), study suchw.c. If you want to modify trace
amplitudes in a single-trace process (Section 10.6), we suggest you study sugain.c.

C-15
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

This page has been intentionally left blank

 Appendix D: Makefiles: Alternative to Shell Scripts

D. Makefiles: Alternative to Shell Scripts

by Chris Liner
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

D.1 Introduction
Makefiles are commonly used for software installation on Unix systems. But make
can be used for other purposes as well. As a small example, I will describe a makefile
(found in Section D.4) that reproduces the functionality of the shell script model1.sh
(Section 4.2).
First, some general comments about working with makefiles. There can be only one
makefile in a directory and, for our purposes, it will always be named “makefile.” It is a
text file like any other and can therefore be edited by whatever tools you use to edit shell
scripts. I use vi, which is a simple editor built into every Unix system. To speed things
up. I define an alias in my environment called vimake:
alias vimake 'vi makefile'
There are variations from system to system, but this kind of alias is usually defined in
a “dot” file, such as .login or .cshrc (talk to your local system expert for advice). With
this alias in place, you can navigate to the working directory you want, then type vimake
and be editing the makefile.
A makefile can perform a simple or complex series of interrelated tasks. For example,
the Stanford Exploration Project at Stanford University uses make to generate processing
results, figures for publication, and even entire books. To be specific, at this time they use
a variant called gmake, but I will describe generic make here.

D.2 Discuss the Makefile

Before describing what is inside the makefile, let's just run it. In what follows, % is
the Unix prompt. The first thing to try is:
% make
In a well-designed makefile, this will do nothing except echo some information. The
last thing you want is a dangerous makefile sitting there waiting for you to come back in
3 months and casually enter make only to have it remove or modify files that cannot be
recovered. Enough on safe computing.
For our makefile, typing:
% make
invokes a default response which is just a little text explaining what is available via this
makefile:
Available Makes:
model1.dat ps clean
Example: make model1.dat
This tells us we have three make choices. The first creates model data, the second
creates a postscript file showing the model, and the third cleans the current working
directory.
Let's look inside the makefile now. The first things we see are some parameters that
will be used by later make items. Anything with a # in the first location is a comment

D-1
 Appendix D: Makefiles: Alternative to Shell Scripts

ignored by make, but very useful for anyone who wants to understand your makefile
(including you at some later date). So the lines:
# Experiment Number
num=1
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

give a comment and then a value is assigned to the parameter named num. Note that in
the actual makefile these lines are flush left. Makefiles are touchy about alignments and
spaces versus tabs, so it is always a good idea to copy a working makefile and then
modify it to your needs; otherwise, a lot of time can be wasted trying to figure out
formatting problems. Parameter num now has the value 1, and to use it somewhere later
we need to refer to it in a particular way. For example we can use $(num) or ${num}; for
our purposes these are equivalent. The next four lines of the makefile include two more
comments and two parameter assignments:
# Name output binary model file
modfile=model${num}.dat
# Name output encapsulated Postscript image file
psfile=model${num}.eps
Notice that, as in shell scripts, the parameter assignments can be embedded, meaning
that parameters can be set based on previously set parameters. Parameters modfile and
psfile are thus defined in terms of the num parameter.
So much for parameter assignments; now on to the main action. The default action of
the makefile is described by:
default:
@echo " "
@echo "Available Makes:"
@echo " model${num}.dat ps clean"
@echo " Example: make model${num}.dat"
@echo " "
Again, it is very important that “default:” begin in column 1 and end with a colon,
and equally important that all lines in the default routine begin with a tab, not a sequence
of spaces. The makefile interprets these lines in this way: “When the user types make
with no following characters, I will go to the first executable step, which happens to be
named default in this case, and it tells me to echo 5 lines of text back to the command
line, then stop.”
When we enter make, we again see:
Available Makes:
model1.dat ps clean
Example: make model1.dat
verifying that parameters like ${num} have been replaced with their values, thus
generating an echo like model1.dat from model${num}.dat.
The next make item actually gets something done:
${modfile}:
rm -f ${psfile}
trimodel xmin=0 xmax=6 zmin=0 zmax=2 \
xedge=0,6 \
.
.
.
sfill=0.1,1.5,0,0,0.25,0,0 \

D-2
 Appendix D: Makefiles: Alternative to Shell Scripts

> ${modfile}
The name of this item is given by parameter modfile, which in turn depends on
parameter num. In fact, it is not easy to tell what we type to invoke this item. But from
earlier typing of make, we know it is:
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

% make model1.dat
that will get the job done.
The first action is to remove the postscript file whose name is stored in parameter
psfile. Next, SU program trimodel is executed with several lines of parameters (not all
are shown here), and the resulting model file is created with a name given by parameter
modfile. As with SU shell scripts, each continuation line must end in a “\”,not a space. To
get secondary indentation (good style), use a tab and 2 spaces. The number of spaces is
up to you, but the tab is necessary or the makefile will not work.
The next item in our makefile is invoked by:
% make ps
This executes the following make code:
ps: ${modfile}
# Create a Postscipt file of the model
# Set gtri=1.0 to see sloth triangle edges
spsplot < ${modfile} \
gedge=0.5 gtri=2.0 gmin=0 gmax=1 \
title="Earth Model - 5 layers [M${num}]" \
labelz="Depth (km)" labelx="Distance (km)" \
dxnum=1.0 dznum=0.5 wbox=6 hbox=2 \
> ${psfile} \
Note the name of this item is hardwired to “ps” and that something sits on the name
line to the right of the colon (:). Make interprets this line this way: “To make ps, I need a
file called model1.dat. If it exists in the current directory, then I will proceede to execute
this item. If it does not exist, then I will make it before continuing.” This is an example
of a conditional make. To make ps, we need to make model1.dat, unless a file of that
name is already in the directory. This is a powerful idea. For example, a finite difference
simulation might take 24 hours to create data that we want to display. With a conditional
make item, the simulation will not be redone unless the data file is missing. If the data
file is there, then only the display will be done.
For this to work right, make items have to be strictly matched to their dependencies.
Consider the following example make items:
data:
suspike ... > data.su
ps: data
supsimage .... > fig.ps
If we enter “make ps”, then make thinks it needs a file named “data.” Not finding it,
make will run the make item “data” and generate a data.su file. There is still no file
named “data,” but make is happy it did something logical and it goes on to create the
image. So this is not a conditional make, but an absolute one. Invoking “make ps” will
always make new data. The way to make this conditional is:
data.su:
suspike ... > data.su
ps: data.su

D-3
 Appendix D: Makefiles: Alternative to Shell Scripts

supsimage .... > fig.ps

Notice that the name of the “data.su” make item is matched to the file name it
creates, and both are matched to the file need of the “ps” item. Now if we invoke “make
ps” it will only remake data.su if it is not present in the local directory.
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

D.3 “clean” and Final Thoughts

The last make item has no dependencies. It simply removes unwanted files.
clean:
rm *.dat *.eps
It is good practice to design makefiles so that “make clean” leaves only the makefile
and necessary files in the directory. By “necessary,” I mean permanent data files (like a
SEG-Y archived version) and that is about it. Everything else should be generated by the
makefile. Also, if you are working on a publication, make items can be keyed to your
document. For example, “make fig11” might generate fig11 from scratch. This goes a
long way toward the goal of reproducible research, and also lets other people quickly
adapt your work to their own data and problems.

D.4 The Makefile

# Experiment Number
num=1
# Name output binary model file
modfile=model${num}.dat
# Name output encapsulated Postscript image file
psfile=model${num}.eps

default:
@echo " "
@echo "Available Makes:"
@echo " model${num}.dat ps clean"
@echo " Example: make model${num}.dat"
@echo " "

${modfile}:
rm -f ${psfile}
trimodel xmin=0 xmax=6 zmin=0 zmax=2 \
xedge=0,6 \
zedge=0,0 \
sedge=0,0 \
xedge=0,2,4,6 \
zedge=0.30,0.50,0.20,0.30 \
sedge=0,0,0,0 \
xedge=0,2,4,6 \
zedge=0.55,0.75,0.45,0.55 \
sedge=0,0,0,0 \
xedge=0,2,4,6 \
zedge=0.65,0.85,0.55,0.65 \
sedge=0,0,0,0 \
xedge=0,2,4,6 \
zedge=1.30,1.30,1.60,1.20 \
sedge=0,0,0,0 \
xedge=0,6 \
zedge=2,2 \

D-4
 Appendix D: Makefiles: Alternative to Shell Scripts

sedge=0,0 \
kedge=1,2,3,4,5,6 \
sfill=0.1,0.1,0,0,0.44,0,0 \
sfill=0.1,0.4,0,0,0.40,0,0 \
sfill=0.1,0.6,0,0,0.35,0,0 \
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

sfill=0.1,1.0,0,0,0.30,0,0 \
sfill=0.1,1.5,0,0,0.25,0,0 \
> ${modfile}

ps: ${modfile}
# Create a Postscipt file of the model
# Set gtri=1.0 to see sloth triangle edges
spsplot < ${modfile} \
gedge=0.5 gtri=2.0 gmin=0 gmax=1 \
title="Earth Model - 5 layers [M${num}]" \
labelz="Depth (km)" labelx="Distance (km)" \
dxnum=1.0 dznum=0.5 wbox=6 hbox=2 \
> ${psfile} \

clean:
rm *.dat *.eps

D.5 The Makefile Edited

Unfortunately, we cannot usually see tabs, just their effect on spacing. Since tabs are
an essential part of makefiles, the makefile of Section D.4 is presented again below. This
time, invisible tabs are replaced by <TAB> and line numbers are added for discussion.
Notice the unexpected tab on line 15. Do not overlook the “>” on lines 42 and 52.
1 # Experiment Number
2 num=1
3 # Name output binary model file
4 modfile=model${num}.dat
5 # Name output encapsulated Postscript image file
6 psfile=model${num}.eps
7
8 default:
9 <TAB>@echo " "
10 <TAB>@echo "Available Makes:"
11 <TAB>@echo " model${num}.dat ps clean"
12 <TAB>@echo " Example: make model${num}.dat"
13 <TAB>@echo " "
14
15 ${modfile}:<TAB>
16 <TAB>rm -f ${psfile}
17 <TAB>trimodel xmin=0 xmax=6 zmin=0 zmax=2 \
18 <TAB> xedge=0,6 \
19 <TAB> zedge=0,0 \
20 <TAB> sedge=0,0 \
21 <TAB> xedge=0,2,4,6 \
22 <TAB> zedge=0.30,0.50,0.20,0.30 \
23 <TAB> sedge=0,0,0,0 \
24 <TAB> xedge=0,2,4,6 \
25 <TAB> zedge=0.55,0.75,0.45,0.55 \
26 <TAB> sedge=0,0,0,0 \
27 <TAB> xedge=0,2,4,6 \

D-5
 Appendix D: Makefiles: Alternative to Shell Scripts

28 <TAB> zedge=0.65,0.85,0.55,0.65 \
29 <TAB> sedge=0,0,0,0 \
30 <TAB> xedge=0,2,4,6 \
31 <TAB> zedge=1.30,1.30,1.60,1.20 \
32 <TAB> sedge=0,0,0,0 \
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

33 <TAB> xedge=0,6 \
34 <TAB> zedge=2,2 \
35 <TAB> sedge=0,0 \
36 <TAB> kedge=1,2,3,4,5,6 \
37 <TAB> sfill=0.1,0.1,0,0,0.44,0,0 \
38 <TAB> sfill=0.1,0.4,0,0,0.40,0,0 \
39 <TAB> sfill=0.1,0.6,0,0,0.35,0,0 \
40 <TAB> sfill=0.1,1.0,0,0,0.30,0,0 \
41 <TAB> sfill=0.1,1.5,0,0,0.25,0,0 \
42 <TAB>> ${modfile}
43
44 ps: ${modfile}
45 <TAB># Create a Postscipt file of the model
46 <TAB># Set gtri=1.0 to see sloth triangle edges
47 <TAB>spsplot < ${modfile} \
48 <TAB> gedge=0.5 gtri=2.0 gmin=0 gmax=1 \
49 <TAB> title="Earth Model - 5 layers [M${num}]" \
50 <TAB> labelz="Depth (km)" labelx="Distance (km)" \
51 <TAB> dxnum=1.0 dznum=0.5 wbox=6 hbox=2 \
52 <TAB>> ${psfile} \
53
54 clean:
55 <TAB>rm *.dat *.eps

D-6
 Appendix E: On the CDs

Appendix E: On the CDs

E.1 Category List
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

The following scripts are included:

Ngazmig.sh fxdisp.scr killer2.sh psmerge2a.sh
Niviewcvp.scr fxdisp.sh migStolt.sh psmerge2b.sh
Niviewcvp.sh idecon.scr migcvp.sh psmerge3.sh
Nmigcvp.sh idecon.sh migmovie.sh shotrwd1.sh
Nmigmovie.sh ifk.scr model1.sh shotrwd2.sh
Nnmo.sh ifk.sh model2.sh shotrwd3.sh
Nsort2cmp.sh igain.scr model3.sh shotrwd4.sh
Tmigcvp.sh igain.sh model4.sh showcmp.sh
Tmigmovie.sh iva.scr myplot.sh showshot.sh
Tnmo.sh iva.sh nmo4.sh showshotB.sh
acq1.sh iview.scr nmo4a.sh sort2cmp.sh
acq2.sh iview.sh oz14prep.sh tf.sh
acq3.sh iviewcvp.scr oz14velan.sh tvQC.sh
acq4.sh iviewcvp.sh psmerge1a.sh velanQC.scr
acq4shot.sh killer.sh psmerge1b.sh velanQC.sh

The following ASCII file is included:

killer.txt (output of killer.sh)

The following seismic data files are included:

seis4.su output by acq4.sh (Section 6.3)
27 Mbytes

Nshots.su Nankai: 2-D line of shot gathers (Section 10.1)

424 Mbytes (from Prof. Greg Moore, U. of Hawaii)

Nstack.su Nankai: stack 2-D line (Section 10.1)

27 Mbytes (from Prof. Greg Moore, U. of Hawaii)

Tshot.su Taiwan: 2-D line of shot gathers (Section 12.1)

411 Mbytes (from Prof. Greg Moore, U. of Hawaii)

The following programs are included:

sukeycount.c
sudumptrace.c
tvnmoqc.c

The following makefile is included:

makefile

E-1
 Appendix E: On the CDs

The following figures are included in the “figures” directory:

fig-07.5.tif fig-10.8.top.tif fig-13.6.top.tif
fig-07.6.tif fig-10.8.bottom.tif fig-13.6.bottom.tif
fig-07.8.tif fig-10.9.top.tif fig-13.7.top.tif
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

fig-07.9.tif fig-10.9.bottom.tif fig-13.7.bottom.tif

fig-07.10.tif fig-11.1.left.tif fig-14.1.tif
fig-07.11.tif fig-11.1.right.tif fig-14.2.top.tif
fig-07.12.right.tif fig-11.5.tif fig-14.2.bottom.tif
fig-08.1.tif fig-13.1.top.tif fig-14.3.tif
fig-08.2.tif fig-13.1.bottom.tif fig-14.4.top.tif
fig-08.5.tif fig-13.2.top.tif fig-14.4.bottom.tif
fig-08.6.tif fig-13.2.bottom.tif fig-14.6.tif
fig-09.5.tif fig-13.3.top.tif fig-14.7.tif
fig-10.5.top.tif fig-13.3.bottom.tif fig-14.8.tif
fig-10.5.bottom.tif fig-13.5.top.tif fig-15.2.left.tif
fig-10.7.tif fig-13.5.bottom.tif fig-15.2.right.tif
fig-15.3.tif
The “figures” directory has all the color figures and a few others, all in uncompressed
TIFF format.

E-2
 Appendix E: On the CDs

E.2 Alphabetical List – Unix (excluding figures)

Ngazmig.sh acq4shot.sh makefile seis4.su
Niviewcvp.scr fxdisp.scr migStolt.sh shotrwd1.sh
Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

Niviewcvp.sh fxdisp.sh migcvp.sh shotrwd2.sh

Nmigcvp.sh idecon.scr migmovie.sh shotrwd3.sh
Nmigmovie.sh idecon.sh model1.sh shotrwd4.sh
Nnmo.sh ifk.scr model2.sh showcmp.sh
Nshots.su ifk.sh model3.sh showshot.sh
Nsort2cmp.sh igain.scr model4.sh showshotB.sh
Nstack.su igain.sh myplot.sh sort2cmp.sh
README iva.scr nmo4.sh sudumptrace.c
Tmigcvp.sh iva.sh nmo4a.sh sukeycount.c
Tmigmovie.sh iview.scr oz14prep.sh tf.sh
Tnmo.sh iview.sh oz14velan.sh tvQC.sh
Tshot.su iviewcvp.scr psmerge1a.sh tvnmoqc.c
acq1.sh iviewcvp.sh psmerge1b.sh velanQC.scr
acq2.sh killer.sh psmerge2a.sh velanQC.sh
acq3.sh killer.txt psmerge2b.sh
acq4.sh killer2.sh psmerge3.sh

E.3 Alphabetical List – PC (excluding figures)

acq1.sh iviewcvp.sh nmo4.sh shotrwd4.sh
acq2.sh killer.sh nmo4a.sh showcmp.sh
acq3.sh killer.txt Nnmo.sh showshot.sh
acq4.sh killer2.sh Nshots.su showshotB.sh
acq4shot.sh makefile Nsort2cmp.sh sort2cmp.sh
fxdisp.scr migcvp.sh Nstack.su sudumptrace.c
fxdisp.sh migmovie.sh oz14prep.sh sukeycount.c
idecon.scr migStolt.sh oz14velan.sh tf.sh
idecon.sh model1.sh psmerge1a.sh Tmigcvp.sh
ifk.scr model2.sh psmerge1b.sh Tmigmovie.sh
ifk.sh model3.sh psmerge2a.sh Tnmo.sh
igain.scr model4.sh psmerge2b.sh Tshot.su
igain.sh myplot.sh psmerge3.sh tvnmoqc.c
iva.scr Ngazmig.sh README tvQC.sh
iva.sh Niviewcvp.scr seis4.su velanQC.scr
iview.scr Niviewcvp.sh shotrwd1.sh velanQC.sh
iview.sh Nmigcvp.sh shotrwd2.sh
iviewcvp.scr Nmigmovie.sh shotrwd3.sh

E-3
 Seismic Data Processing with Seismic Un*x: A 2D Seismic Data Processing Primer
Supplementary material

Supplementary material for this book can be accessed at

Downloaded 06/03/14 to 134.153.184.170. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/

http://link.aip.org/link/mm/doi=10.1190/1. 9781560801948.supp&filename=262E_supp.zip