EARTH-IMAGER

content
• Initial settings
• Read a data file
• File management and file extension
• Edit data file
• Inversion settings
• Run inversion
• View graphics
• Copy image to clipboard
• Save image
• Save data in XYZ format
• Save STG data
Initial settings
• First look at EarthImager
• The typical data processing procedures with EarthImager:
• 1, Read a data file
• 2, Change inversion settings or read a settings file
• 3, data editing statistics (optional)
• 4, Run inversion,
• 5, View data misfit histogram to remove data with a high data misfit
(optional)
• 6, Run inversion again (optional)
• 7, View Inverted Resistivity Section,
• 8, Change image properties such as min/max contour level, vertical
exaggeration (phóng đại) factor,
• 9, print the inversion results (images),
• 10, Save the inversion results (Images and ASCII output in an XYZ format)
 Initial settings
1,Click settings/initial settings, then Initial settings
window pops up.
Mean:
+ Specify what data to reject at the initial data
loading,
+ what inversion algorithm to use,
+ the function of Y coordinates (depth vs line offset),
the orientation of the vertical axis, vertical electrode
alignment,
+ and whether to save the processing results or not.

+ Definition of Y axis and orientation of vertical axis

must be set before reading the data file for them to
them to take effect. Other initial settings may be set
before or after reading data file.
Initial settings
• There are seven criteria for removal of noisy data on the Initial Settings window.
• Any data beyond these thresholds will be flagged as noisy data, You may also
manually delete noisy data by using mouse clicks and the Delete key.

• For surface data, the flagged noisy data will be shown in black dots. But for
borehole data, the flagged noisy data will disappear . These flagged noisy
data are not really removed before starting the inversion.
• If you change any criterion for noisy data removal from the Initial Settings
Window, all manual data edits and electrode removal will be lost.
Therefore, you have to redo these edits.
• A singular surface data point, defined as a single low or high apparent resistivity
value at a certain depth level of the pseudo-section, will be removed by default.
• When the user clicks the Start Inversion button, a clean data set is saved in the
current trial folder and is reloaded back for inversion.
• If you do not want to save anything, you should uncheck “Save Output”.
 Read a data file
• There are two ways to load a data file:
+ choose the Read Data menu item from the File menu or
+ click the first tool button .
Two coordinates are used to define electrode locations
- The first coordinates X is the relative location of an electrode along the survey line.
The second coordinate Y (or Z) is used in EarthImager to define the depth of an
electrode for cross borehole resistivity tomography.
Format: o AGI Sting and SuperSting data files (*.stg), EarthImager can also read
popular
DAT files, URF files, some IPR-12 files and some GDP-32 TDIP data files
+ Universal Resistivity (data) File (URF) format to facilitate processing of resistivity
data collected with non-AGI instrument
Read a data file – Universal resistivity data (URF)
Read a data file – Universal
resistivity data (URF)
 Read a data file
• After reading data successfully, EarthImager will display either an apparent
resistivity pseudosection for surface survey or an apparent resistivity scatter
plot for borehole Electrical Resistivity Tomography (ERT) data.
Read a data file – ERT data
File management and file extension
File management and file extension – A trial folder
Edit data file
• There are three ways to identify and remove noisy data

+ Some possibly bad electrodes would be singled out for removal

+ Based on the initial data trimming threshold on the Initial Settings window. The
noisy data will be automatically removed based on the threshold.

+ A user would identify and remove noisy data manually with the mouse clicks
and Delete key.
Edit electrodes
• helps identify and remove bad electrodes. All data associated with the bad
electrode will be removed. Both for ERT and surface data.
• Each electrode is assigned a vote calculated from user-specified criteria.
• The higher vote an electrode has, the noisier the electrode is. An electrode
with a vote larger than half of the number of electrodes is a suspected bad
electrode. A horizontal or vertical linear feature on ERT scatter plot is usually
a sign of bad electrode.
•
Edit electrodes
• To identify possibly bad electrodes in the electrode editor,
1) may change any “voting” criterion by clicking the Settings button. This will
lead to the Initial Settings page of the Inversion Settings window.

2) Click the second column header Number of Noisy Data to sort this column at
a descending order so that the “noisiest” electrodes are emerged at the top of
the list.

3) Check the box in front of the electrode to be removed. Uncheck the box if
you change your mind before clicking the Remove button. However, the
electrode removal is final and can not be undone. If a good electrode is
removed, you may reload the data file and start from the beginning.

4) Click the Remove button to remove the selected electrodes.

Edit electrodes
Edit surface data
Edit ERT data
Shift and reverse profile
• Choose Edit| shift and Reverse profile menu item to launch the shift and
Reverse Profile window.
Removal of poorly-fit data
• At the end of the inversion, display a data misfit histogram by choosing view |
convergence and data misfit| Data misfit Histogram.
Scale distance
• the electrode coordinates are not scaled appropriately due to an operational
error in the field or during the creation of a command file.
Data editing statistics
Debug surface raw data
Debug surface raw data
View contact resistance
• Contact resistance is an excellent indicator of data quality and it can also help
identify bad electrodes. A contact resistance value of over 10,000 Ohms is a
clear sign of a poorly installed electrode which often leads to unacceptable
data quality. An acceptable contact resistance value is often below 5000
Ohms, ideally below 2000 Ohms.
Inversion settings
 Run inversion
• Start inversion by green tool button:

• Menu Inversion| Start Inversion

4. INVERSION SETTINGS EXPLAINED
• All inversion control parameters are arranged on Settings Window. There
are four pages of inversion settings: Initial Settings, Forward Model,
Resistivity Inversion, and IP Inversion. You may go to desired settings page
directly from Settings menu or click Inversion Settings tool button to bring
up Settings Window. The Default Combo Box at the bottom of Settings
Window will reset all settings to hard-coded default settings for several
typical scenarios such as Surface, Cross-Borehole ERT, CRP-Freshwater,
and CRP-Saltwater. The User Settings option in the Default Combo Box
corresponds to the settings of the last inversion. Other three buttons at
the bottom follow the Windows convention
Resistivity inversion settings
Resistivity inversion settings
Survey Planning Process
• Two-step process:
+ 1st, a synthetic data is created based on your assumption of subsurface
resistivity distribution and electrode configuration, a command file. > this step
is a virtual survey and also called forward simulation.
+ 2nd, the synthetic data is inverted to reconstruct the subsurface resistivity
distribution. The assumption of resistivity distribution and the inverted
resistivity section will be displayed on the screen. > called inverse simulation.
The goals of our survey planner are to help:
1, find the optimal data collection strategy
2, find the optimal forward modeling settings
3, find the optimal inversion settings
4, find the limit of resistivity imaging, that is, what is resistivity contrast, what
target size, what target depth and/or what kind of subsurface geology the
resistivity imaging can resolve.
Forward Simulation
• Run Forward simulation:
• 1, Read command file| the file menu or go to next step > will load an AGI
command file (*.cmd) and create a mesh based on the initial settings after
successful loading of the command file or a model file.
• If electrode spacing is not 1m or 1ft, should enter a scale factor other than 1.0
here.
Forward Simulation
• 2, Read Model File from the File menu.
Will load an AGI model file (*.mdl).
• A model file: A model file consists of
measurement commands, a synthetic
model, and mesh settings.
• 3, choose Forward modeling settings
from the settings menu and change the
settings as you desire.
• 4, the Model settings panel would
appear to the right of the main
window.
• To hide this panel, right-click the
mouse on the panel and choose the
only option “Hide model settings
panel”
• 5, If borehole electrodes are detected, the Survey Planner would enter the
Electrical Resistivity Tomography (ERT) mode automatically. A typical ERT
mesh may look like this.
21, Forward simulation from the Tool menu > will start the forward
simulation.
• 22, For borehole resistivity imaging
• 23, To format the graphics, choose tool buttons on the top toolbar under
the main menu.
24) To browse graphics at different views, choose desired menu item from
View menu: Synthetic Model, and Synthetic Model and Data.
25) You may save or print your on-screen image through the File menu or
tool buttons.
26) A synthetic model file (*.mdl) is saved in the current trial folder at the
end of forward simulation. You may also save the model file by choosing
the menu File |
Save Synthetic Model.
Survey Planner
• The survey planner is a combined forward and inverse simulation.
• First, a virtual survey is carried out by forward modeling (simulation) to
generate a synthetic data set.
• The synthetic data set is inverted to reconstruct the subsurface model which is
compared against the synthetic model for resolution analysis.
• To run survey planner steps for forward simulation:
• 1, Choose your inversion method from the Initial Settings page of the Settings
Window.
• 2, Set your desired inversion settings from the Resistivity Inversion settings
page of the Settings Window.
• 3, If IP inversion is planned, choose appropriate IP inversion settings. This
should be done before loading the command file.
Survey Planner
• 4, The estimated noise percentage and stopping RMS on resistivity and IP
inversion settings windows should be always equal to or greater than the
random noise added to synthetic data.
5, Instead of choosing Forward Simulation, choose Survey Planner from the
Tools menu.
6) You will see the forward simulation graphics first, and then 6) You will see the
forward simulation graphics first, and then inversion images.

7, For the surface resistivity surveys, three images are shown. The top one is the
synthetic apparent resistivity (or chargeability) data pseudosection. The middle
one is the inverted resistivity (or chargeability) section. The bottom one is the
synthetic resistivity (or chargeability) model. It is important to compare the
synthetic model with the reconstructed model and to understand the resolution
of resistivity/IP imaging. You should note that it is impossible to recover the
exact subsurface resistivity distribution. Instead, you get a smooth model after
the inversion.
Survey Planner
Survey Planner
• 8, For IP survey planning, inverted resistivity and chargeability sections are
shown at the end of IP inverse simulation.
• 9, For borehole resistivity tomography, the synthetic resistivity model and
the reconstructed model are shown side by side for comparison.
• 10) To format the graphics, choose tool buttons on the top toolbar under the
main menu.
11) To browse graphics at different views, choose the desired menu item
from the View menu: Synthetic Model, Synthetic Model and Data, and Survey
Planner Results. The survey planner results are images described in the steps
7) and 9).
12) For IP survey planning, one may click the Resistivity or IP tab at upper left
corner of the main image window to toggle between the resistivity and IP
images of the same layout.
• 13) You may save the onscreen image in a bitmap file.
14) You may save the inverted resistivity section in a text file in the XYZ format
Survey Planner
5.4 A-Priori Model
• While building a synthetic model, one may set part of the model as a-priori
model which will remain unchanged during inversion of synthetic data. By
constraining part of the model with known information, one has a better
chance to resolve the unknown part of a synthetic model.
Check the box “A-Priori Model” (see above) before picking/selecting model
blocks. The model blocks selected after checking the “A-Priori Model” box will
be held constant during the simulation. Be sure to uncheck the box if the
following model blocks are not going to be set as apriori model. In the example
above, a 10 Ohm-m conductive block is set as a-priori model, otherwise, the
survey planner would not be able to resolve it so accurately. The a-priori model
status can be saved in the synthetic model file and read back into EarthImager
2D.
•
5.5 Survey Planner with Underwater Electrodes
• This case is the same as the borehole electrodes. The electrodes can be
either underwater or underground. One should set the water layer as the a-
priori model by using the feature described in the last section.
Set negative Z coordinates to indicate the depth of electrodes in the
electrode geometry file. There is a sample GEO file (underwater.geo) and a
corresponding model file (underwater.mdl) in the demo folder. In this
model, the top blue layer is set as a-priori model.
5.6 Survey Planner with Topography
• This feature helps simulate resistivity surveys with topography.
Electrodes in a command file are assumed to be laid out on a flat ground
surface. Elevation measurements are listed in a terrain file. It is important
to maintain an agreement of distance measurement format between the
command file and the terrain file. This is the same requirement as that in
topographical data inversion described in the section 6.1. The screenshot
below is a synthetic scenario and it demonstrates how the survey planner
with topography looks like.
•
5.6 Survey Planner with Topography
• 1, Read a command file, for example, Dip56R8.cmd in the demo folder. Use a
scale factor 3 for this SuperSting command file. A mesh with flat ground
surface would show up.
• 2, Read a terrain file, for example, pyramid.trn in the demo folder. The mesh
would be transformed to fit the topography. If the mesh does not change,
you may want to change the default settings to Surface and reload both
command file and terrain file.
• 3, Build the synthetic model as usual.
• 4) Start either Forward Simulation or Survey Planner.
View graphics - toolbar
Input graphics title
Input Project Information
Raw resistivity data
Inverted resistivity section
Pseudo and inverted resistivity section
Convergence curve
Data misfit pseudosection
Data misfit crossplot
Borehole data misfit scatter plot
Model sensitivity
• This shows relative model sensitivity. To obtain relative model sensitivity,
first, multiply transposed sensitivity matrix by sensitivity matrix (JTJ), and
then extract its diagonals.
• This diagonal has the length of the model vector. The relative sensitivity
value should be always greater than zero.
Time-lapse difference image
Synthetic model
• This shows the mesh and model of the forward simulation or the survey
planner. This is the interface to build a synthetic resistivity model. For
further information, refer to the Survey Planner section.
 Synthetic model and data
• This shows the apparent resistivity pseudosection of forward simulation
with the synthetic model at the bottom. You may modify the model at
the bottom and re-run the forward simulation or the survey planner
from this interface. For further information, refer to the Survey Planner
section.
 Survey planner results
• This shows the survey planner result. The survey planner is, in fact, a
forward and inversion simulator. On the screen, you will see three
images. The synthetic data from numerical simulation is shown at the
top. The middle section is the inverted resistivity model. The synthetic
resistivity model is shown at the bottom. You may modify the model at
the bottom and re-run the survey planner from this interface. For
further information, refer to the Survey Planner section.
Data misfit histogram
• At the end of inversion, use data misfit histogram to remove poorly fit
data. See Section 3.5.5 for
detail
Continuous Resistivity Profile (CRP)
• CRP makes
processing of
extremely large data
sets possible and
efficient. CRP inverts
a large data set
section by section
and assembles a
long inverted
resistivity section at
the end of inversion.
Reciprocal data error scatter plot
Copy image to clipboard
Save image
Save data in XYZ format
Save STG data
Save synthetic model
• There are two types of model files in
EarthImager.
• An MDL file is a combined command
file and synthetic model tied to the
modeling mesh. An MDL file can be
read into EarthImager for forward
and inverse simulation.
• An MOD file lists a-priori resistivity
and IP model blocks and it is used to
constrain the inversion. An MOD file
can be created manually in a text
editor or in EarthImager. Here is a
sample MOD file.
Read A-priori model

• Go to the menu File | Read A-Priori Model to input an MOD file to

constrain the inversion.
• EarthImager will try to keep model blocks in the MOD file unaltered
during the iterative inversion process.
• An MOD file can be created manually in a text editor or in EarthImager
Survey Planner. Refer to the MOD file format in the last section.
6.1 Inversion with Terrain
• Resistivity survey with topography possesses two difficulties. First, electrode
positioning on topographic surface is more difficult than that on flat ground
surface. Second, inversion of resistivity data with topography is challenging.
EarthImager has a highly optimized finite element algorithm so your inversion
with topography will be extremely fast and reliable. By the way, we provided a
flexible format of terrain files, so it is easier for you to position the electrodes
in the field.
6.1.1 Terrain file format
• EarthImager has no restrictions on the electrode spacing and array
configuration. Electrodes may be placed anywhere you prefer. You have two
ways to tell EarthImager your electrode geometry. First, you may use
horizontal distance that is equivalent to the horizontal coordinate. Second,
you may use tape measure or slope distance whose projection on the flat
surface is the horizontal coordinate. Here is a sample terrain file.
• The terrain file (*.trn) is a comma separated text file that may be created in
any text editor such
as the NotePad.
Run inversion with terrain
• 1, Click the Read Data button, e.g., Pyramid.stg.
2) Modify inversion settings if you desire. Default surface settings are a safe
option for surface data.
3) Click File/Read Terrain File, e.g., Pyramid.trn.
4) Click the green Start Inversion button.
Inversion of resistivity data with topography uses the finite element modeling
method because the triangular mesh of finite element method is very flexible
to cope with topographical variations. The sensitivity and inverted resistivity
sections will show variations of surface
elevation. The vertical axis is labeled with the elevation instead of the depth
in a standard inversion without terrain. Borehole data inversion with
topography was supported from the version 1.7.7, but there is no field ERT
data with terrain in our release. To demonstrate borehole data inversion with
terrain correction, the user may invert ERT3bh.stg with the tweaked terrain
file ERT3bh.trn.
Input A-Priori Information
• 1) Click the Read Data button.
2) Modify inversion settings if you desire.
3) Click Settings/Input A-Priori Information menu to display the GUI interface
to build a synthetic model.
4) Follow the instruction in Section 5.2 about how to build a synthetic model.
5) Click the green Start Inversion button.
A-priori resistivity value deviated from the true subsurface resistivity greatly
would slow down the inversion process and impose artifacts on the inverted
resistivity section. The inversion process would NOT retain an erroneous a-
priori resistivity value.
6.3 Time Lapse Inversion
• The optional Time Lapse Inversion module is used for in-situ monitoring applications. It could
be, for example, used to detect leakage at a landfill site, a dam, or underground storage tank. In
such a case, electrodes are installed at fixed locations during the period of monitoring to facilitate
image comparison and high sensitivity of small changes in the ground. First, a base (or baseline,
background) resistivity data set is collected, so a base resistivity model of the site can be
established as a reference model. Then, the monitor survey is repeated during the period of
monitoring with the same command file as the one used in the base survey. Our Time Lapse
Inversion algorithm takes advantage of the base survey data and base resistivity model. Instead
of inverting the monitor data set alone, our algorithm inverts the difference between the monitor
and base data sets. Moreover, the base resistivity model is used as the a-priori model in the time
lapse inversion. Therefore, our time lapse inversion converges faster than standard inversion and
is more sensitive to small subsurface changes. The coherent inversion artifacts may be canceled
in the difference images.
6.3 Time Lapse Inversion
• A time lapse monitoring application may have one base dataset and one or
more monitor datasets. The time lapse inversion produces difference images
between any monitor dataset and the base data set.
The result of time lapse inversion is presented as the percent difference
between the two sections. The time lapse method may be used for
monitoring of any subsurface processes. Some common situations include
potential leakage from landfills, dams and underground storage tanks. Other
monitoring situations include fracture tracing by injection of a conductive
solution such as a salt solution, saltwater intrusion in coastal areas,
environmental remediation progress, groundwater recharge, infiltration
studies, and so on.
The time lapse inversion may be carried out in two different ways depending
on the type of input. Difference inversion accepts two raw data files (STG) as
inputs. This is a comparison of before and after images. Time lapse inversion
requires the output of base data inversion and
inverts a list of monitor data sets in a batch mode.
6.3.1 Difference inversion
• Difference inversion combines the inversion of the base dataset and time lapse
inversion of a
single monitor dataset in one step. It is applicable to the cases with one base
dataset and a single
monitor dataset. Here are the steps to follow for a difference inversion.
1) Start EarthImager 2D. A dongle option code for time lapse inversion must be
activated in
order to proceed with the difference inversion.
2) Choose the menu Inversion | Time Lapse Inversion and then the “Difference
Inversion”
tab.
3) Input the base data file by clicking the corresponding Browse button. The demo
data set
is C:\Program Files\AGI\EarthImager2D\demo\TimeLapse0.stg.
4) Input the monitor data file by clicking the corresponding Browse button. The
demo data
set is C:\Program Files\AGI\EarthImager2D\demo\TimeLapse1.stg.
 6.3.1 Difference inversion
• 5, Input a terrain or underwater terrain file if any.
6) Choose “Same Command File” or “Different Command Files”. It is strongly
recommended that you use the same command file for the base data and monitor
data
collections.
7) Change the inversion settings by clicking the “Change Settings” button.
8) Start Difference Inversion. EarthImager will invert the base dataset first and then
do a
difference inversion of the monitor dataset. The inverted results at the optimal
iteration
based on the RMS error and/or L2-norm will be saved in the output (OUT) file, which
may be loaded back into EarthImager for custom settings and visualization.
9) At the end of difference inversion, percent difference of resistivity image is
shown. It is
preferred to set the absolute value of minimum and maximum plot bounds to be
equal, so zero (no change) will be in the center of the color scale and in a green
color by default.
10) In addition to all images for standard data inversion, we provided two more
images for time lapse inversion: percent change of resistivity and percent change of
conductivity, under the menu item View | Time Lapse Images
6.3.2 Time lapse inversion
• The difference inversion scheme in the last section requires inversion of the base
dataset for every monitor dataset. This turns out to be unrealistic for a long term
monitoring application that usually has many monitor datasets. So we developed a
time lapse inversion scheme that can handle a large number of monitor datasets in
a batch mode and output an AVI movie file at the end of the batch inversion.
• In our time lapse inversion scheme, the base dataset must be inverted first like a
standard 2D data inversion. We suggest that the user runs the inversion of base
dataset two or more times to clean up the base dataset with the histogram.
However, excessive data removal is discouraged because any data point removed
from the base dataset will also be removed from all monitor datasets. The
inversion output file of the base dataset is served as the input of the time lapse
inversion. All monitor datasets are inverted with the base inversion results as the a-
priori information. To start the time lapse inversion, follow these steps.
6.3.2 Time lapse inversion
• 1, Choose the Time Lapse Inversion menu item from the Inversion menu and
select the Time Lapse Inversion page. Make sure that the base dataset has
already been inverted.
• 2, Input the Base Inversion Output (*.out) File by clicking the Browse button.
This file consists of previously inverted results of the base dataset. The base
inversion output is served as the a-priori information of the time lapse
inversion. While adding the OUT file to the input box, EarthImager checks the
availability of the inversion settings file in the same folder of the OUT file. The
settings file will be used for the time lapse inversion. It is preferred that the
same settings are used for both base data inversion and monitor data
inversion.
6.3.2 Time lapse inversion
• 3, if a terrain (*.TRN) or underwater terrain (*.UWT) file exists in the base
inversion output folder, it will be used in the time lapse inversion.
4) Add monitor data files to the list. There are two ways to add monitor data file
names. First, the user may click the “Add Monitor Data Files” button and select
multiple data files from the Windows File Open Dialog. Second, an ASCII batch
file with a list of (STG) data file names and, if any, the base output file name
may be loaded into EarthImager 3D. A batch file is a simple list of file names
with a complete file path. If an OUT file is found in the batch file, it will
overwrite the OUT file already in the input box. But the monitor data files will
be simply appended to the current monitor file list without any check on
duplicates. Here is a sample time lapse batch file:
• C:\Program
Files\AGI\EarthImager2D\demo\TimeLapse0\trial1\TimeLapse0_trial1.out
C:\Program Files\AGI\EarthImager2D\demo\TimeLapse1.stg
C:\Program Files\AGI\EarthImager2D\demo\TimeLapse2.stg
C:\Program Files\AGI\EarthImager2D\demo\TimeLapse3.stg
6.3.2 Time lapse inversion
• 5, To delete a monitor file from the file list, select the file and then press the Delete key. To
change the order of a monitor file, use the Up and Down arrow to move it up or down.
All monitor files must be listed in a sequential order based on the data collection time
because an AVI movie will be created at the end of inversion to show time lapse changes.
Page 84 of 139 Instruction Manual of AGI EarthImager 2D, 7/6/2009
6) Save the batch file for future use.
7) The user may change the inversion settings by going to the Main Window and choosing
any settings page, but this is strongly discouraged.
8) An AVI movie file may be produced with the difference images from all monitor datasets
when three or more monitor datasets are inverted in a batch mode. To create an AVI file,
all difference images must be presented with the same color scale and data range (Min
and Max).
9) Set Min Change and Max Change based on the above discussion and convention. It is
going to be difficult to set these two numbers without any prior knowledge. We
recommend that you set Min-Change = -1 * Max-Change. This will help suppress some
artifacts and make the green color no-change zones.
6.3.2 Time lapse inversion
• The selection of the plotting Parameter will depend on the application type. For a case
with increasing conductivity, it is almost mandatory to choose “Conductivity Change” as
the parameter to plot. For a case with increasing resistivity, select “Resistivity Change”
as the plotting parameter. Choice of Min Change, Max Change and Parameter is not fatal
but an inappropriate selection will fail the attempt of AVI movie creation. The time lapse
inversion results can be read back into EarthImager for visualization and data analysis.
11) Click the Start button to start the time lapse inversion in a batch mode. The monitor
data files are inverted one by one.
12) At the end of inversion of each monitor data set, the inverted resistivity data is saved as
images in the JPG and BMP formats and ASCII XYZ format. Moreover, both resistivity
difference and conductivity difference between the monitor model and base model are
saved in the bitmap format and ASCII XYZ format. Here are some typical output files
from a time lapse inversion of one monitor dataset. Bitmap files are used to create an
AVI movie to demonstrate the animated time lapse change of the subsurface. Go the
Create AVI Movie under the Tools menu for detail. Three images below demonstrate
gradual decrease of the subsurface resistivity from synthetic data sets.
• 13, The size of a saved image would have the same size of the image
rendering window.
14) A time lapse inversion output file may be read back into
EarthImager one at a time for
visualization and further data analysis. This will be discussed in the
next section.
15) At the end of inversion, an AVI movie file is created and played
automatically if three or
more monitor data sets are inverted in a batch mode.
6.3.3 Read time lapse output file
• To load the time-lapse inversion results back into EarthImager, go to the
menu item File | Read Time Lapse Output File or the menu Inversion | Time
Lapse Inversion and then choose the page “Read Time Lapse Output File”.
• Trial directories for both base data inversion and monitor data inversion are
required to produce the percent difference images. The user is simply
prompted to choose or enter these two directories. Note that the inversion
settings (INI) files in these two directories must be intact for successful
display of difference images.
6.4 Inversion of ERT Data
• When you look at the CAVE.STG and AMISTAD.STG, you may find that abs(V/I) is in the
range of 10-2 ~ 100. ERT data is usually noisier due to the large dynamic range of signal,
i.e., V/I varies from 10-6 ~ 100. These small V/I's are usually extremely hard to fit and
produce a large RMS. So RMS may be a misleading measure to show ERT data misfit. Here
are some tips for ERT data inversion.
• • Remove Noisy Data When Reading the STG File
1, Minimum abs(V/I). Change settings before reading the file. The data rejection threshold
is on the Initial Settings window. Min V/I should be set between 10-4 and 10-3. 10-4 is a
too conservative setting. Any V/I data smaller than 10-3 are usually hard to fit and
produce very large RMS. But you are not supposed to remove all of them. The Suppress
Noisy Data setting on Resistivity Inversion settings window will give less respect to
hard-to-fit data.
2) Negative apparent resistivity is common in ERT data. You are not supposed to remove all
of them. Negative apparent resistivity data are always shown in black dots in the ERT
data scatter plot. There are often many troublemakers that have a negative apparent
resistivity. The horizontal or vertical linear feature of black dots is the sign of bad
electrodes.
3) Apparent resistivity threshold should be set based on the common sense.
6.4 Inversion of ERT Data
• Edit ERT Data
Edit Electrodes. If you see any horizontal or vertical linear feature of black
dots on the ERT data scatter plot, go to the Electrode Editor under the Edit
menu and remove the bad electrode. By the way, all data associated with the
bad electrode will be removed at once.
2) Edit ERT Data. After you read the ERT data file, the ERT data scatter plot is
the interface to edit the data one by one. Select the noisy data by mouse
clicking and remove it with the Delete key. Noisy data may be defined as a
very small or very large apparent resistivity, or a negative apparent resistivity,
or any standout data point on the scatter plot. You may repeat this on the
data misfit scatter plot after the inversion is done, removing poorly-fit data.
6.4 Inversion of ERT Data
• • Resistivity Inversion Settings
1) Choose a stabilizing factor >= 100 because noisy ERT data usually produce
a rough model. A larger stabilizing factor leads to a smoother model.
2) Choose a Resolution Factor >= 0.20, e.g., 0.25, to enhance the resolution
in the central part of two boreholes.
3) Choose a Horizontal / Vertical Roughness ratio <= 0.5 to force a layered
earth in the inverted model. This may exacerbate the borehole or electrode
effects.
4) Check ‘Suppress Noisy Data’. This will treat hard-to-fit data with less
respect. The stop criterion “L2-Norm” will be automatically checked. The
inversion will stop when the normalized L2-norm reaches 1.
6.6 Batch Inversion
• Batch inversion processes many data files in a batch mode without
interruption. Its input is a batch file (*.bch) containing a list of data file names
and inversion settings file names. Here is an example of a batch file:
• C:\Program Files\AGI\EarthImager2D\demo\cave.ini
C:\Program Files\AGI\EarthImager2D\demo\Amistad.stg
C:\Program Files\AGI\EarthImager2D\demo\cave.stg
C:\Program Files\AGI\EarthImager2D\demo\ERT3bh.ini
C:\Program Files\AGI\EarthImager2D\demo\ERT3bh.stg
C:\Program Files\AGI\EarthImager2D\demo\IPtest1.stg
C:\Engineering\Projects\pc\Volente\DemoData\terrain.ini
C:\Engineering\Projects\pc\Volente\DemoData\terrain.stg
C:\Engineering\Projects\pc\Volente\DemoData\terrain.trn
Batch Inversion
• EarthImager reads these files one by one. When EarthImager reads an
INI settings file, it updates the inversion settings and uses these settings
for inversion of the next data file. When EarthImager reads a data file
(*.stg, *.dat), it starts inversion of this data set, and saves the inverted
resistivity section, and IP section if available, as a JPEG image file and a
text data file in the XYZ format. The chosen iteration has a minimum L2
norm if Suppress Noisy Data is checked or has a minimum RMS error. If
EarthImager detects a terrain file (*.trn) immediately follows a data file,
it will carry out the inversion with terrain correction. Any file with a file
extension other than INI, STG, DAT and TRN will be ignored.
Incompatible data files will be skipped. Any surface data file with one or
more reference (infinity or remote) electrodes may be skipped if it has a
non-zero Y coordinate for the infinity electrode.
• To start batch inversion, choose the Batch Inversion menu item from the
Inversion menu. The batch inversion interface will show up.
the steps to carry out the batch inversion
• 1, Load a batch file (*.bch). A batch file is a text file containing a list of file names. See the
example batch file above. You may create a batch file in any text editor such as the
NotePad or in this batch inversion interface.
2) Add data and/or settings files to the list. You may add more data and settings files to the
list even if you already loaded a batch file. To select and add more than one files at a
time, use the mouse button and the CTRL or SHIFT key.
3) Remove files added by mistake. Select the file and press the Delete key. You may also
select many files at once by holding down the mouse button and drag. Then a single
Delete key press will delete many files.
4) You may move one or more files up or down using the Move Up or Move Down button.
To move a file, you need to select it first. This function is helpful when the INI settings
file is not at the right position or the terrain file (*.TRN) does not immediately follow the
corresponding data file.
5, After all adding and editing, now you may save the file list as a batch file (*.bch) so you
may load it back.
6) To save inverted section in an XYZ format without two triangular edges, check the
“Exclude Edges” checkbox. For more information about this, refer to the Section 3.11.
the steps to carry out the batch inversion
• 7, To add a border and a project information table to the graphics, click the
Project Settings button.
8) Start Batch Inversion. The data files in the list will be inverted one by one
and you may monitor the inversion progress on the screen. Inverted resistivity
section (and IP section if applicable) at the iteration with the minimum RMS
error will be saved as a text file in the XYZ format and a JPEG graphics file. All
inversion results are also saved in the OUT file that may be loaded back to
EarthImager for analysis.
• Note that user interactions are skipped in the batch mode. Therefore, you can
not edit the data and customize your graphics. In this case, you may want to
set more strict data rejection thresholds.
6.8 Continuous Resistivity Profiler
• The Continuous Resistivity Profiler (CRP) module (optional) in EarthImager 2D
processes the resistivity and IP data collected with a boat-towed array, or a
pulled-array, or a roll-along array. The primary feature of these types of data
sets is a large number of dynamic electrodes, say, tens of thousands of
electrodes, due to the continuous movement of the array. The processing
(inversion) of a data file with a large number of electrodes is computer memory
demanding and computation-intensive. A marine resistivity survey with a boat-
towed array produces thousands of measurements with thousands of dynamic
electrodes in an hour. It is almost impossible for a standard inversion program
to process such a large number of data. We developed the Continuous
Resistivity Profiler (CRP) module to attack this difficult problem.
6.8 Continuous Resistivity Profiler
• The CRP module uses our unique Divide-and-Conquer strategy to deal with a large
number of electrodes. When a data file with a large number of electrodes is loaded,
EarthImager divides the long section into many subsections and inverts subsections
one after another. At the end of the inversion, a complete inverted section is
assembled and shown in the main window for acomplete and compressed image and
a scrollable window for a detailed view.
• The CRP module can invert both resistivity and IP data simultaneously. It can also
process a large roll-along data file with terrain correction. From the version 1.9.0, the
user may choose one of three axis label formats: linear distance, longitude/latitude,
and UTM coordinates. When the mouse cursor hovers on the inverted resistivity
image, the location and resistivity value would beshown on the bottom status bar.
• What makes the CPR module impressive and attractive is that all complicated details
of Divideand-Conquer are done behind the scene. To a user, inverting a large data set
with tens of thousands of electrodes is almost the same as inverting our demo Sting
Cave data set with 28 electrodes. A user simply reads the large data set, and then
Start Inversion. This simple two-step process is very user-friendly
1) Read STG data file
• EarthImager will find out if your data should be processed in the CRP mode.
CRP is an optional module of EarthImager and it is not included in the basic
version of EarthImager. If the number of electrodes or dynamic electrode
locations in your data is over a threshold, say 200 electrodes for marine data,
EarthImager would suggest CRP to process your data. If your dongle
(hardware key) was not activated for CRP, a message dialog shown below will
pop up, and your data will be inverted in a standard inversion method. This
can be painfully slow and the inversion may terminate if you have a low
computer memory (RAM). The demo CRP data file is CRP.stg, a marine data
set with over 2000 dynamic electrode locations. The corresponding water
depth file is CRP.dep.
1) Read STG data file
• If your data set consists of too many electrodes and your dongle was activated
for CRP, EarthImager will automatically enter the CRP processing mode. There is
a CRP check box on the Settings | Initial Settings window for deactivating the
CRP mode. If this CRP check box is unchecked and you read in a large data file
with many electrodes (say, over 200), a dialog box would pop up to ask you to
confirm:
1) Read STG data file
• After reading a large data file, EarthImager splits the long section into
overlapped smaller sections and creates a batch inversion process. The user
may view the file list by choosing the menu item Inversion | Batch Inversion.
Deleting any STG file(s) at the end of the spreadsheet is allowed at this stage. It
is not appropriate to delete any file in the middle of the file list.
2) Read water depth file
• An AGI SuperSting marine system collects not only electrical resistivity data,
but also water depth and water temperature data simultaneously. Our Marine
Log Manager (MLM) preprocesses these data and outputs an STG resistivity
data file and a depth file consisting of both water depth and temperature
information.

Water depth data is read into EarthImager for display and implementation of
an a-priori model. The temperature data is displayed below the inverted
resistivity section for comparison and correlation analysis.
• Choose the menu File| Read Water Depth File to load a water depth file
(*.dep). A complete water depth file consists of water depth, average water
resistivity and water temperature data. A Water depth file is usually generated
by AGI Marine Log Manager (MLM). Here are the first few lines of demo depth
file CRP.dep
Read a water depth file

The water depth file helps (a) draw a white water depth line in the inverted resistivity sections
(b), set water resistivity as a-priori model, and (c) display a water temperature profile for joint data
interpretation.
Starting from the version 1.9.0, EarthImager 2D accepts an upgraded water depth file with 5 columns of data
generated by the AGI marine log manager (MLM) version 1.3.3.182 or newer
Here are top few lines of a new sample depth file.
Fix water resistivity
• The resistivity of water layer is determined from one of these three approaches.
First, if the water resistivity is given in the water depth file, the water layer will be
set with a single constant resistivity value. Second, when the water resistivity value
is NOT given in the water depth file, apparent resistivity values estimated with closet
pairs of transmitting and receiving electrodes are used to define the water
resistivity. Apparent resistivity values vary along the profile. So this approach gives a
water layer with varying resistivities. However, this approach will fail if a gradient
array is used because apparent resistivity values from a gradient may be greatly
affected by the water bottom sediments. The last approach to set water resistivity is
to input a water conductivity file obtained from water conductivity measurements.
Refer to the descriptions below for detail.
• Three approaches described above help determine the resistivity of water layer,
which sets the starting model of CRP data inversion. To fix the resistivity of water
layer, one must check the “Fix Water Resistivity” checkbox on the CRP settings page.
If the convergence is not achieved in five iterations, the resistivity of water layer is
set free for any changes.
Read terrain file
• For a large roll-along data set with topography, one should read a terrain file
(*.trn) after reading the STG data file. EarthImager will automatically insert
the terrain file into the batch process.
Read CRP water conductivity file
• If you have a water conductivity meter and take water conductivity
measurements while collecting resistivity data, the water conductivity
measurements may be used to constrain the CRP inversion. The top water
layer would hold on given water conductivity values without changes during
the iterative inversion. To read a water conductivity file, go to File|Read CRP
Water Conductivity File. Here are top few lines of a water conductivity file
(*.CON).
• Note that the time stamp (both date and time) in the conductivity file must
be synchronized with the time stamp in the AGI sting (STG) data file.
EarthImager depends on the correct time stamp to figure out the
measurement location of the conductivity meter.
Change inversion settings
• Default settings by choosing from the Default Combo Box at the bottom of the
settings window should be a safe option. We hard-coded several sets of default
settings for various typical scenarios such as surface, cross-Borehole ERT, CRP-
Freshwater, and CRP-Saltwater. The user settings option in the Default Combo Box
corresponds to the settings of the last inversion. Here are a few settings of
importance of CRP data inversion.
1, Minimum Voltage in mV: Marine data signal can be very low. This number must
be to below 0.1 mV. 0.02 mV and 0.05 mV would be good options
2, Minimum abs(V/I) (ohm): The recommended setting would be 1.0e-05
3, Min Apparent Res (ohm-m): Seawater resistivity is around 0.25 ohm-m, and
fresh water resistivity may range from 10 ohm-m to 30 ohm. Therefore, 0.05 ohm-
m and 0.1 ohm-m should be good options for seawater, and 1 ohm-m and 5 ohm-m
are good settings for fresh water
4, Number of Mesh Division: set it to 1 to speed up the inversion.
5, Min Resistivity (ohm-m): This setting is similar to Min Apparent Res
6, Model Parameter Width: The default is 4. It should be equal to or greater than
2. This is a horizontal smoothing factor.
7) Start inversion

• Just like doing the standard inversion, click the green arrow tool-button or the
Start Inversion menu. EarthImager will take care of everything. The inversion
will proceed from the left (smaller X coordinate) to the right (larger X
coordinate) subsection by subsection. The inversion progress can be observed
from the title caption at the top of the main Window. The Screen shot below
shows EarthImager is inverting the second one of the total 16 subsections.
• For each subsection, EarthImager saves these six files: an inversion settings file
(*.ini), a clean STG file after removal of some noisy data, an inversion output
file (*.out), a JPG file of inverted resistivity section at the iteration with the
smallest RMS error, a data file (*.dat) of inverted resistivities in the XYZ format,
and a blanking file (*.bln) for compatibility with the surfer graphics package.
The OUT file may be loaded back into EarthImager for further study.
• At the end of CRP inversion, the complete inverted resistivity section is
shown in the main window in a contracted form, and a long scrollable
section is displayed in another window as a
detailed view. These images are saved in the current trial folder as JPEG
files. The inverted data is saved in an XYZ format for visualization in other
graphics programs. If the longitude and latitude data is available in the
water depth file, the inverted resistivity data is also saved in an LLT
(long/lat) file AND a UTM file. An LLT file consists of four columns with the
first column as the longitude coordinate in degrees, the second column as
the latitude coordinate in degrees, the third column as the depth in meters
or feet, and the last column as the inverted resistivity value in ohm-m. The
first two columns of a UTM file are Easting (X) and Northing (Y) in meters.