Wave Analysis

Short Term and Long Term Analysis

Haryo Dwito Armono. PhD

Updated 4 March 2009
Introduction
 Short term analysis:

www.aviso.oseanobs.com
 analyse of wave that
occurs within one
wave train
 Long term analysis
 derivation of statistical 0.04

distributions that cover 0.03

0.02
Water Level (m)
many years 0.01

-0.01

-0.02

-0.03

-0.04
0 5 10 15 20 25 30
Time (sec)
 0.04

0.03

0.02
Water Level (m)

0.01

-0.01

-0.02

-0.03

-0.04
0 5 10 15 20 25 30
Time (sec)
Wind – Wave correlation
Significant wave height and
wind speed are shown in
July 2007, units are metres
and metres per second
respectively. These figures
highlight the relationship
between wind speed and
significant wave height: the
faster the wind, the highest
the waves.The lowest waves
(dark blue) are mainly in
tropical and subtropical
oceans, where the smallest
wind speed are recorded.
www.aviso.oceanobs.com/en/applications/atmosphere-
wind-and-waves/wind-and-waves/seasonal-
variations/index.html
Why collect wave data?
 Monitoring of coastal processes
such as beach erosion and
sediment transport.
 Baseline design statistics for
coastal projects.
 Operational assistance in coastal
construction projects.
 Monitoring of severe weather
conditions.
 Oceanographic research.
(Manly Hydraulic Lab : http://mhl.nsw.gov.au/)
How to measured wave?
 Use wave recorder : i.e:
 Wave staffs
 e.g: www.oceansensorsystem.com
 Wave rider / wave buoy
 e.g: www.datawell.nl
 Pressure sensor
 e.g : www.civiltek.com
 Satellite images
 e.g : GFO (Geosat Follow On)
 Place wave recorder in deepwater (>0.5L)
 Record wave height, period and direction (duration 15’ – 60’)
 Links http://cdip.ucsd.edu, http://mhl.nsw.gov.au/,
http://www.coastal.udel.edu/coastal/comps.html, etc
 Assignment 2 :
 Find more info on wave recorder (product detail, vendors, measurement and analysis
methods, etc)
Wave Animation
 www.datawell.nl

Typical wave recorder

TRITON-ADV
www.sontek.com

www.civiltek.com

www.oceansensorsystem.com

AWAC
www.nortek.com
http://www.scienceprog.com/wp-content/uploads/2007i/Ocean_embedded/wave_heigh_measurement.jpg
 Typical recorded samples
File created by: WAVELOG on 12/11/2006 1:53:28 AM
Device: Model 730W_CTP
Serial No: 17528
File Type: WAVE STATS
Operating Mode: Wave Burst
Contract Reference: None
Site Reference: None
Location Latitude: None
Location Longitude: None
Setup by: None
User Info: NO SITE INFORMATION
Pressure Calibration: P = 1:RPT200-1791,3500,35,958161,1,040897
Temperature Calibration: T = 12 +3.7484500e-10 +1.7482800e-03 +1.0768100e-01
Conductivity Calibration: C = 13 +3.8447599e-13 -9.6605406e-09 +3.0098500e-03 -7.8249698e-02
File Number: 0
Sample Rate: 0.4
Density: 1025.97
Gravity: 9.807
Instrument Height: 0
Trigger Level: 0
Wave Data Save Trigger Level: 0
Tide Burst Duration (secs): 600
Tide Cycle Time (secs): 1200
Wave Burst Duration (secs): 5119.99992370606
Wave Cycle Time (secs): 1200
Speed of sound formula Chen & Millero
Pressure formula: 1
Date Time Hs Tp Hmax E Tp
m(U) Secs m(U) N.m(U)^2.m^-3 Time Secs Hs Hmax
9/12/2006 12/9/06 23:16 0.0134 0.0210 0.1125224 12/9/06 23:16 0.0134 0.0210
9/12/2006 12/9/06 23:36 0.0224 0.0352 0.3157163 12/9/06 23:36 0.0224 0.0352
9/12/2006 23:56:15 0.0139 0.0218 0.1213013 12/9/06 23:56 0.0139 0.0218
10/12/2006 0:16:15 0.0117 0.0184 8.65E-02 12/10/06 0:16 0.0117 0.0184
10/12/2006 0:36:15 19647.13 meragukan 12/10/06 0:56 0.0165 0.0258
10/12/2006 0:56:15 0.0165 4.413793 0.0258 0.170476 12/10/06 1:16 4.413793 0.0168 0.0263
10/12/2006 1:16:15 0.0168 4.129032 0.0263 0.1766452 12/10/06 1:56 4.129032 0.0144 0.0226
10/12/2006 1:36:15 4.129032 11155.82 meragukan 12/10/06 2:16 4.129032 0.0125 0.0196
10/12/2006 1:56:15 0.0144 4.129032 0.0226 0.1302911 12/10/06 2:36 4.129032 0.0046 0.0073
10/12/2006 2:16:15 0.0125 4.129032 0.0196 9.84E-02 12/10/06 2:56 4.129032 0.0136 0.0213
10/12/2006 2:36:15 0.0046 0.0073 1.35E-02 12/10/06 3:16 0.0134 0.0210
10/12/2006 2:56:15 0.0136 4.129032 0.0213 0.1155402 12/10/06 3:36 4.129032 0.0128 0.0201
10/12/2006 3:16:15 0.0134 3.657143 0.0210 0.1127914 12/10/06 3:56 3.657143 0.0215 0.0338
10/12/2006 3:36:15 0.0128 3.878788 0.0201 0.1033529 12/10/06 4:56 3.878788 0.0111 0.0174
10/12/2006 3:56:15 0.0215 42.66667 0.0338 0.2907289 12/10/06 5:16 4.266667 0.0111 0.0174
10/12/2006 4:56:15 0.0111 3.878788 0.0174 7.75E-02 12/10/06 5:36 3.878788 0.0109 0.0171
10/12/2006 5:16:15 0.0111 3.878788 0.0174 7.71E-02 12/10/06 5:56 3.878788 0.0099 0.0156
Triton Webinar
 Typical Analysis Results

www.sontek.com
www.datawell.nl
Terms 0 .0 4
Ensemble of Three Realizations

 Realization 0 .0 3
1

W a te r L e v e l (m )
0 .0 2

0 .0 1

 Representative record at 0

-0 .0 1
particular time range -0 .0 2

-0 .0 3

 Ensemble -0 .0 4
0 5 10 15 20 25 30
T im e (s e c )

 Compilation of several
realization 0 .0 4
0 .0 3

W a te r L e v e l (m )
2
 Each ensemble has 0 .0 2
0 .0 1

parameters such as mean, 0

-0 .0 1

standard deviation, -0 .0 2
-0 .0 3

skewness, kurtosis, etc -0 .0 4

0 5 10 15 20 25 30
T im e (s e c )

0 .0 4
0 .0 3
W a te r L e v e l (m )

0 .0 2 3
0 .0 1
0
-0 .0 1
-0 .0 2
-0 .0 3
-0 .0 4
0 5 10 15 20 25 30
T im e (s e c )
Terms 0 .0 4

0 .0 3
a

W a te r L e v e l (m )
0 .0 2

0 .0 1

 Stationary -0 .0 1
0

 If none of the ensembles -0 .0 2

-0 .0 3

parameter (z) vary in time -0 .0 4

m2a = m2b = m2c

0 5 10 15 20 25 30
exp: T im e (s e c )

sa = s b = sc
m3a = m3b = m3c 0 .0 4
0 .0 3
b

W a te r L e v e l (m )
m4a = m4b = m4c 0 .0 2
0 .0 1
0
-0 .0 1

 ErgodicTime -0 .0 2
-0 .0 3

zk ( a )  zk ( b )  zk ( c ) 
-0 .0 4
0 5 10 15 20 25 30

zk 
T im e (s e c )


3 
K N  harus sama 0 .0 4

zk  j  
1 1

0 .0 3
W a te r L e v e l (m )

c
zk  0 .0 2

K N k 1 j 1  0 .0 1
0
-0 .0 1
-0 .0 2
-0 .0 3

Assume ergodicity in wave records -0 .0 4

0 5 10 15 20 25 30

Kamphuis, Intro to Coastal Eng & Management T im e (s e c )

Short Term Analysis
 Height of waves () are random
 It is impossible to predict exact value of  at any
time or location
 Probability that  has a certain value is called
PDF (Probability Density Function), p().
 p() can be described by normal distribution.

1   2  1 N
p( )  exp  2  , s 2   2j
s 2  2s  N j 1
 t t R N
  2j
1 1
s 2   2 = lim t R 
tR  t 0  2 dt 
N j 1

Short Term Analysis cont’d

 Assumed that H = 2max, then PDF for H

(Probability that H has a certain value)

H2  H 2 
 p( H )  exp  2  Rayleigh Distribution
4s 2
 8s 
N
1
s   j
2 2

N j 1
Short Term Analysis cont’d

 The Cumulative Distribution Function (CDF) of wave

height: probability that any individual wave height H’ is less
than a specified wave height H
H
H2  H 2   H 2 
 P( H '  H )   2 exp  2  dH  1  exp  2 
0
4s  8s   8s 
 The Probability of Exceedance : the probability that any
individual wave of height H’ is greater than a specified
wave height H :

 H 2 
 Q( H '  H )  1  P( H '  H )  exp  2 
 8s 
 Annual Directional Percentage Occurence and
Omnidirectional Percentage Exceedance of Significant Wave Height

% %
N NE E SE S SW W NW
Total Exceed
Total % 15,51 5,90 6,85 5,82 10,33 13,41 19,79 22,39 100
6.0 - 6.5 0,01 0,01 0,01 0,01 0,01 0,05 0,05
5.5 - 6.0 0,01 0,02 0,06 0,10 0,02 0,02 0,23 0,28
5.0 - 5.5 0,05 0,05 0,02 0,03 0,15 0,06 0,05 0,12 0,53 0,81
4.5 - 5.0 0,10 0,03 0,05 0,05 0,11 0,04 0,20 0,44 1,02 1,83
4.0 - 4.5 0,09 0,04 0,09 0,12 0,09 0,29 0,71 1,43 3,26
3.5 - 4.0 0,24 0,02 0,06 0,12 0,21 0,44 0,83 1,92 5,18
3.0 - 3.5 0,47 0,07 0,11 0,13 0,29 0,30 0,72 1,45 3,54 8,72
2.5 - 3.0 0,63 0,15 0,29 0,36 0,76 0,63 1,36 1,80 5,98 14,7
2.0 - 2.5 1,84 0,34 0,34 0,57 1,37 1,88 3,07 3,31 12,72 27,42
1.5 - 2.0 3,91 1,17 1,22 1,06 2,77 2,94 4,55 4,95 22,57 49,99
1.0 - 1.5 6,13 2,71 2,63 2,02 3,17 4,34 5,69 5,98 32,67 82,66
0.5 - 1.0 1,96 1,32 1,96 1,32 1,23 2,81 3,21 2,74 16,55 99,21
0.0 - 0.5 0,08 0,05 0,14 0,06 0,13 0,09 0,18 0,06 0,79 100
 Short Term Analysis cont’d

1
0.9
P
0.8
Cumulative
0.7 Distribution Function
(H).p(H), P, Q

0.6 Rayleigh Distribution

0.5
(H).p(H)
0.4
0.3
Q
0.2
0.1
Probability of
Exceedance
0
0 1 2 3 4 5 6 7 8
H/Sigma
Short Term Analysis cont’d

 Wave height with Probability of Exceedance Q:

1
 H Q  8s ( ln Q)  2s 2ln  
2

Q

N
1
s 
2

N
 j
 2

j 1
Short Term Analysis cont’d
Short Term Analysis cont’d
 Example 3.1 Calculation of Short-Term Wave Heights

 To analyze a wave record it must be stationary. Hence, it is normal

to record waves for relatively short time durations (10 to 20
minutes). A longer record would not be stationary because wind
and water level variations would change the waves. Thus it is usual
to record, for example, 15 minutes every three hours. It is
subsequently assumed that the 15 min. record is representative of
the complete three hour recording interval.

 Suppose the analysis of such a record yields

T  10 sec and s  1.0 m

 We want to calculate significant wave height Hs, average wave

height H , average of the highest 1% of the waves H 0.01, and the
maximum wave height in the record.
Short Term Analysis cont’d
 Practice :
 - see Queens CIVL 456 Tutorial No 4
Long Term Analysis
 Purpose :
 To organise wave height data
 To extrapolate data set to extreme values of
wave height occuring at low probability of
exceedance
 Methods
 Statistical
Analysis of grouped wave data
 Extreme Value Analysis from ordered data
Grouped wave data
4.5
(4)
4 Nov 01 - Nov 30, 1983
(2) (5)
3.5
Wave Height (m)

3 (1)
Ht = 1.5 m
2.5
(3)
2 (7) (8)
(6)
1.5
1
0.5
0
0 100 200 300 400 500 600 700 800
Time (hrs from Nov 01, 0:00, 1983)
Grouped wave data cont’d

Data in the left table obtained

from 34,9 years of record.
 = number of data points / yr
= 282306 / 34.9 = 8089
= 2738 / 34.9 = 78.45
Grouped wave data cont’d
P  P( H '  H )
1019
P( H '  1.75)   0.372
2738
1019  549
P( H '  2.00)   0.573
2738
1019  549  382
P( H '  2.25)   0.712
2738
Q  Q( H '  H )  1  P

Curveline is difficult to interpolate!  transformed into straightline

Curve Transformation Normal Distribution
Cumulative Distribution Function

 Normal Probability Distribution

 Transferred to:
 Log-Normal Probability Log Normal Distribution

Distribution
Gumbel Distribution

 Gumbel Distribution
Weibul Distribution

 Weibull Distribution

y=A.x + B
Distribution Models

sH = standard deviation,
H = mean wave height
 and b = Weibull and Gumbel Parameter
g = lower limit of H = threshold value in a Peak over Threshold data set
Probability Table

Normal Distribution Curve z = -3.4  P = 3.37 x 10-4

z = +3.4 P =1 - 3.37 x 10-4
= 0.999663
P = 3.37 x 10-4 In Excel : NORMINV

z = -3.4
z =0
 (1) (2) (3) (4) (5) (6) (7) (8) (9)
H N P Q z ln H G W W
=0.8 =1.3
1.75 1019 0.372 0.628 -0.326 0.560 0.012 0.384 0.555
2.00 549 0.573 0.427 0.183 0.693 0.584 0.816 0.883
2.25 382 0.712 0.288 0.560 0.811 1.081 1.316 1.184
2.50 254 0.805 0.195 0.859 0.916 1.528 1.848 1.459
2.75 174 0.869 0.131 1.119 1.012 1.959 2.421 1.723
3.00 113 0.910 0.090 1.339 1.099 2.359 2.996 1.964
3.25 81 0.939 0.061 1.550 1.179 2.772 3.627 2.210
3.50 60 0.961 0.039 1.766 1.253 3.232 4.366 2.477
3.75 40 0.976 0.024 1.976 1.322 3.713 5.176 2.750
4.00 27 0.986 0.014 2.190 1.386 4.244 6.105 3.044
4.25 19 0.993 0.007 2.442 1.447 4.916 7.326 3.406
4.50 10 0.996 0.004 2.683 1.504 5.611 8.638 3.769
4.75 4 0.998 0.002 2.849 1.558 6.122 9.632 4.031
5.00 2 0.99854 0.00146 2.976 1.609 6.528 10.436 4.234
5.25 1 0.99890 0.00110 3.063 1.658 6.816 11.014 4.377
5.50 2 0.99963 0.00037 3.378 1.705 7.915 13.276 4.910
5.75 0 0.99963 0.00037 3.378 1.749 7.915 13.276 4.910
6.00 1 1.00000 0.000
Total 2738

X in Distribution Models : [1], [6]

Y in Distribution Models : [5], [7], [8], [9]
Extrapolation
 The Exceedence Probability of one event in TR
years :
Extreme Value Analysis
 If only few major events are known

 Limited number of extreme events

Method
 Rank the data in decreasing order
 Calculate Probability of Exceedence (Q)
i  c1
Q
N  c2
i = ranking of the data point
N = total number of points
 c1, c2 = constants for unbiased plotting position
 Calculate Probability (P)
 Calculate Reduced Variate (z, W, G)
Constant for Unbiased Plotting

Distribution c1 c2
Normal 0.375 0.375
Log Normal 0.250 0.125
Gumbel 0.440 0.120
Weibull 0.2 + 0.27/ 0.2 + 0.23/
Distribution Models

sH = standard deviation,
H = mean wave height
 and b = Weibull and Gumbel Parameter
g = lower limit of H = threshold value in a Peak over Threshold data set
Probability Table

z = -3.4  P = 3.37 x 10-4 Normal Distribution Curve

z = +3.4  P = 1- 3.37 x 10-4
= 0.999663

P = 3.37 x 10-4
In Excel : NORMINV
Example: =NORMIV(J38,0,1)
z = -3.4
z =0
 1
i  c1  1 
Q W   ln 
N  c2  Q
i H Q W i H Q W
1 5.95 0.010 6.675 23 4.22 0.505 0.621
2 5.38 0.033 4.642 24 4.21 0.527 0.572
3 5.26 0.055 3.775 25 4.20 0.550 0.526
4 5.03 0.078 3.227 26 4.20 0.572 0.482
5 4.82 0.100 2.832 27 4.17 0.595 0.441
6 4.75 0.123 2.524 28 4.17 0.617 0.402
7 4.71 0.145 2.274 29 4.16 0.640 0.365
8 4.68 0.168 2.064 30 4.16 0.662 0.330
9 4.63 0.190 1.884 31 4.14 0.685 0.297
10 4.54 0.213 1.727 32 4.14 0.707 0.266
11 4.49 0.235 1.588 33 4.13 0.730 0.236
12 4.43 0.258 1.463 34 4.09 0.752 0.208
13 4.40 0.280 1.351 35 4.09 0.775 0.182
14 4.38 0.303 1.250 36 4.08 0.797 0.156
15 4.36 0.325 1.157 37 4.07 0.820 0.133
16 4.35 0.348 1.071 38 4.07 0.842 0.111
17 4.34 0.370 0.993 39 4.06 0.865 0.090
18 4.33 0.393 0.920 40 4.05 0.887 0.071
19 4.29 0.415 0.852 41 4.04 0.910 0.053
20 4.25 0.437 0.788 42 4.04 0.932 0.036
21 4.24 0.460 0.729 43 4.03 0.954 0.022
22 4.23 0.482 0.673 44 4.01 0.977 0.009
Plot Extreme Distribution
8.0
y=A.x + B
Weibull Reduced Variate - W

7.0
A = 3.395
6.0
B = - 13.704
5.0 y = 3.395x - 13.704
R2 = 0.995
4.0
3.0
2.0
1.0
0.0
-1.0
4 4.5 5 5.5 6 6.5
Wave Height - H (m)

Weibull Distribution for Ordered Data Set (=0.8).

Plot Extreme Distribution
2.5

2.0
y=A.x + B
A = 5.270
Reduced Variate - z

1.5
B = - 7.180
1.0
y = 5.270x - 7.180
0.5 R2 = 0.979

0.0

-0.5

-1.0

-1.5

-2.0
1.4 1.5 1.6 1.7 1.8 1.9
ln H

Log-Normal Distribution for Ordered Data Set

Plot Extreme Distribution
5.0
y=A.x + B
Gumbel Reduced Variate - G

4.0 A = 2.211
y = 2.211x - 8.591
R2 = 0.992 B = - 8.591
3.0

2.0

1.0

0.0

-1.0

-2.0
4 4.5 5 5.5 6 6.5
Wave Height - H (m)

Gumbel Distribution for Ordered Data Set

Distribution Models

sH = standard deviation,
H = mean wave height
 and b = Weibull and Gumbel Parameter
g = lower limit of H = threshold value in a Peak over Threshold data set
Return Period Prediction
Weibull
Return Period (Yrs)
N   b g 20 50 100 200
44 1.26 0.80 0.29 3.97 5.22 5.68 6.05 6.43

Gumbel Return Period (Yrs)

N  b g 20 50 100 200
44 1.26 0.45 3.87 5.31 5.73 6.04 6.36

Log Normal Return Period (Yrs)

N  ln H s 20 50 100 200
44 1.26 1.36 0.19 5.44 5.86 6.16 6.45