Auxiliary materials to the Manual of the LOTOS code with

THE DESCRIPTION OF SEVERAL TESTING EXAMPLES

created using LOTOS-09 version
www.ivan-art.com/science/LOTOS_09

Table of content:

0. Introduction 2
1. DATASET2 (Chile, offshore, synthetic dataset) 2
1.1. DATASET2/model_01 real data inversion, nodes, Vp-Vs scheme 2
1.2. DATASET2/MOD_BL_1 real data inversion, cells, Vp-Vs scheme 4
1.3. DATASET2/BRD_30km synthetic modeling, checkerboard 5
1.4. DATASET2/VBRDmod2 synthetic modeling, vertical resolution 6
1.5. DATASET2/SMILE_01 synthetic modeling, horizontal anomalies 7
1.6. DATASET2/VER_free synthetic modeling, vertical anomalies 8
2. BOLIVAR_ (Costa-Rica, on- and off- shore, synthetic dataset) 9
2.1. BOLIVAR_/model_01 real data inversion, nodes, Vp-Vs scheme 10
2.2. BOLIVAR_/BLOCK_01 real data inversion, cells, Vp-Vs scheme 11
2.3. BOLIVAR_/Ver_BRD1 synthetic modeling, vertical resolution 12
3. TOBA_001 (Toba Caldera, N. Sumatra, synthetic dataset) 13
3.1. TOBA_001/model_01 real data inversion, nodes, Vp-Vs scheme 14
4. TOBA_REA (Toba Caldera, N. Sumatra, real data) 16
4.1. TOBA_REA/real_VPX real data inversion, Vp-Vp/Vs scheme 16
5. Tests for the GAP criterion, AREAS: GAP__180, GAP__280, 18
GAP__360
5.1. GAP__180/model_01 Java Area, events with GAP<180 18
5.2. GAP__280/model_01 Java Area, events with GAP<280 20
5.3. GAP__360/model_01 Java Area, all events within 5 distance 22
6. TURKEY__ (Marmara Sea area, checkerboard) 24
6.1. TURKEY__/model_01 real data inversion, nodes, Vp-Vs scheme 24
7. EGYPT_02 (Egypt area, checkerboard) 25
7.1. EGYPT_02/BOARD_01 real data inversion, nodes, Vp-Vs scheme 25

1
0. Introduction
Here we present several examples of using the LOTOS code for different datasets. In most cases we
use configurations of sources and receivers corresponding to real experiments. However, due to
property right problems we cannot always present observed travel times. Therefore the datasets which
are used as the initial ones are mostly generated artificially by ray tracing through synthetic models.
However, after computing the travel times, we forget about the model and use the same procedure
as for processing real datasets.

All the indicated datasets can be re-run by any person using the presented EXE files for the LOTOS
code which can be freely downloaded from the web page www.ivan-art.com/science/LOTOS_09. If
the parameters are not changed, the results will be the same as shown in this overview. It would be a
good teaching lesson to perform the inversion for the presented examples with different free
parameters (grid spacing, inversion parameters, different starting models etc) to see the influence of
each of them to the result.

When starting processing real datasets, a user can take one of the presented examples as a template.
In this case, one should select a dataset which is the most similar to his/her one. Then one should
change the geographical coordinates of the central point of the study area and prepare two files with
stations and travel times in the inidata folder. Then the inversion can start.

The previews of the obtained results can be seen in folder PICS in the corresponding subfolders.
All the pictures presented here are obtained in this way. At the same time, to produce high-quality
pictures suitable for publication, it is recommended using more sophisticated visualization tools (such
as Surfer).

In case of problems, errors and inconsistencies, please address to the author of the LOTOS code, Ivan
Koulakov ([email protected]).

1. AREA: DATASET2

This dataset is fully artificial. It was designed specially for planning a network in an off-shore
experiment in Central Chile.

Number of stations 12
Number of events 202
Number of P rays: 1952
Number of S rays: 1237

1.1. MODEL: DATASET2/MODEL_01

The processing of this model corresponds to inversion of real data. However, the model has been
created artificially using synthetic model shown in Figure 1 (the model parameters can be seen in
DATASET2/MOD_INI_).

2
Figure 1. Initial model which was used to create the real dataset in
DATASET2/inidata/rays.dat. Picture is from
/PICS/DATASET2/MOD_INI_/hor/syn1_ 2.png

Table 1.1. Parameters of inversion for model DATASET2/model_01

Data processing real
Inversion scheme Vp-Vs
Number of iterations 5
1D optimization YES
Parameterization nodes
Figure with results 2
Approximate calculation time: 35-40 min
Size of the model folder after calculations : 27 Mb

Figure 2. Inversion results for P and S velocity anomalies at 30 km depth based on model
DATASET2/model_01. Pictures are from:
/PICS/DATASET2/model_01/hor/dv15_ 3.png
/PICS/DATASET2/model_01/hor/dv25_ 3.png

3
1.2. MODEL: DATASET2/MOD_BL_1

Same as DATASET2/model_01, but with the use of CELL parameterization. .

Table 1.2. Parameters of inversion for model DATASET2/MOD_BL_1

Data processing real
Inversion scheme Vp-Vs
Number of iterations 5
1D optimization YES
Parameterization CELLS
Figure with results 3
Approximate calculation time: 45-50 min
Size of the model folder after calculations : 37 Mb

Figure 3. Inversion results for P and S velocity anomalies at 30 km depth based on model
DATASET2/MOD_BL_1. Pictures are from:
/PICS/DATASET2/MOD_BL_1/hor/dv15_ 3.png
/PICS/DATASET2/MOD_BL_1/hor/dv25_ 3.png

4
1.3. MODEL: DATASET2/BRD_30km

Checkerboard is represented by alternated cells with horizontal spacing of 30 km and unlimited in the
vertical direction.

Table 1.3. Parameters of inversion for model DATASET2/BRD_30km

Data processing synthetic
Noise: 0.8 of real remnant residuals
Inversion scheme Vp-Vs
Number of iterations 3
1D optimization YES
Parameterization NODES
Figure with results 4
Approximate calculation time: 30 min
Size of the model folder after calculations : 14 Mb

Figure 4. Synthetic model and inversion results for P and S velocity anomalies at 30 km depth based
on model DATASET2/BRD_30km. Pictures are from:
/PICS/DATASET2/BRD_30km/hor/syn1_ 2.png
/PICS/DATASET2/BRD_30km/hor/dv13_ 3.png
/PICS/DATASET2/BRD_30km/hor/dv23_ 3.png

5
1.4. MODEL: DATASET2/VBRDmod2

Checkerboard is represented by alternated cells along the vertical section.

Table 1.4. Parameters of inversion for model DATASET2/VBRDmod2

Data processing synthetic
Noise: 0
Inversion scheme Vp-Vs
Number of iterations 5
1D optimization YES
Parameterization NODES
Figure with model and results 5
Approximate calculation time: 35 min
Size of the model folder after calculations : 16 Mb

Figure 5. Synthetic model and inversion results for P and S velocity anomalies at 30 km depth based
on model DATASET2/VBRDmod2. Pictures are from:
/PICS/DATASET2/VBRDmod2/ver/syn_dv 1.png
/PICS/DATASET2/VBRDmod2/ver/dv15_ 1.png
/PICS/DATASET2/VBRDmod2/ver/dv25_ 1.png

6
1.5. MODEL: DATASET2/SMILE_01

Synthetic test with free anomalies defined in map projection.

Table 1.5. Parameters of inversion for model DATASET2/SMILE_01

Data processing synthetic
Noise: Random: 0.15 s (P), 0.3 s (S)
Inversion scheme Vp-Vs
Number of iterations 5
1D optimization YES
Parameterization NODES
Figure with results 6
Approximate calculation time: 35 min
Size of the model folder after calculations : 15 Mb

Figure 6. Synthetic model and inversion results for P and S velocity anomalies at 30 km depth based
on model DATASET2/SMILE_01. Pictures are from:
/PICS/DATASET2/SMILE_01/hor/syn1_ 2.png
/PICS/DATASET2/SMILE_01/hor/dv13_ 3.png
/PICS/DATASET2/SMILE_01/hor/dv23_ 3.png

7
1.6. MODEL: DATASET2/VER_free

Checkerboard is represented free anomalies defined along the vertical section.

Table 1.6. Parameters of inversion for model DATASET2/VER_free

Data processing synthetic
Noise: 0
Inversion scheme Vp-Vs
Number of iterations 5
1D optimization YES
Parameterization NODES
Figure with model and results 7
Approximate calculation time: 30 min
Size of the model folder after calculations : 15 Mb

Figure 7. Synthetic model and inversion results for P and S velocity anomalies at 30 km depth based
on model DATASET2/VBRDmod2. Pictures are from:
/PICS/DATASET2/VBRDmod2/ver/syn_dv 1.png
/PICS/DATASET2/VBRDmod2/ver/dv15_ 1.png
/PICS/DATASET2/VBRDmod2/ver/dv25_ 1.png

8
2. AREA: BOLIVAR_

This dataset uses the data configuration from a real LET experiment in Costa-Rica. The data in the
inidata folder are computed using the synthetic model which composes the face of Simon
Bolivar (Figure 8). This model remains unchanged with depth. The same model was published in
Koulakov, (2009, BSSA) in the paper which presents the LOTOS code.

Initial dataset:
Number of stations 129
Number of events 1079
Number of P rays: 24285
Number of S rays: 6794
Dataset used for inversion:
Number of stations 108
Number of events 1074
Number of P rays: 24197
Number of S rays: 6744

Figure 8. Initial model which was used to create the real dataset in
BOLIVAR_/inidata/rays.dat. Picture is from
/PICS/BOLIVAR_/TRUE_MOD/hor/syn1_ 1.png

9
2.1. MODEL: BOLIVAR_/MODEL_01

The processing of this model corresponds to inversion of real data. However, the model has been
created artificially using synthetic model shown in Figure 8 (the model parameters can be seen in
BOLIVAR_/MOD_INI_).

Table 2.1. Parameters of inversion for model BOLIVAR_/MODEL_01

Data processing real
Inversion scheme Vp-Vs
Number of iterations 5
1D optimization YES
Parameterization nodes
Figure with results 9
Approximate calculation time: 2 hours 40 min
Size of the model folder after calculations : 68 Mb

Figure 9. Inversion results for P and S velocity anomalies at 20 km depth based on model
BOLIVAR_/MODEL_01. Pictures are from:
/PICS/BOLIVAR_/MODEL_01/hor/dv15_ 1.png
/PICS/BOLIVAR_/MODEL_01/hor/dv25_ 1.png

10
2.2. MODEL: BOLIVAR_/BLOCK_01

Same as BOLIVAR_/MODEL_01, but with the use of CELL parameterization. .

.

Table 2.2. Parameters of inversion for model BOLIVAR_/BLOCK_01

Data processing real
Inversion scheme Vp-Vs
Number of iterations 5
1D optimization YES
Parameterization CELLS
Figure with results 10
Approximate calculation time: 6 hours 30 min
Size of the model folder after calculations : 68 Mb

Figure 10. Inversion results for P and S velocity anomalies at 20 km depth based on model
BOLIVAR_/BLOCK_01. Pictures are from:
/PICS/BOLIVAR_/BLOCK_01/hor/dv15_ 1.png
/PICS/BOLIVAR_/BLOCK_01/hor/dv25_ 1.png

11
2.4. MODEL: BOLIVAR_/Ver_BRD1

Checkerboard is represented by alternated cells along the vertical section.

Table 2.3. Parameters of inversion for model BOLIVAR_/Ver_BRD1

Data processing synthetic
Noise: 0.15 s (P); 0.20 s (S)
Inversion scheme Vp-Vs
Number of iterations 5
1D optimization YES
Parameterization NODES
Figure with model and results 11
Approximate calculation time: 6 hours
Size of the model folder after calculations : 69 Mb

Figure 11. Synthetic model, location of the profile and inversion results for P and S velocity
anomalies at 30 km depth based on model BOLIVAR_/VER_BRD1. Pictures are from:
/PICS/BOLIVAR_/VER_BRD1/ver/syn_dv 1.png
/PICS/BOLIVAR_/VER_BRD1/ver/dv15_ 1.png
/PICS/BOLIVAR_/VER_BRD1/ver/dv25_ 1.png
/PICS/BOLIVAR_/VER_BRD1/ver/map.png

12
3. AREA: TOBA_001

This dataset uses the data configuration from a real LET experiment in the area of Toba Caldera
(N.Sumatra). The data in the inidata folder are computed using the synthetic model which
composes the realistic shapes of anomalies (Figure 12). This model is fully 3D. The same model was
published in Koulakov et al., (2009, JGI).

Initial dataset:
Number of stations 31
Number of events 390
Number of P rays: 3377
Number of S rays: 2462
Dataset used for inversion:
Number of stations 30
Number of events 585
Number of P rays: 3308
Number of S rays: 2413

Figure 12. Initial model of P anomalies which was used to create the real dataset in
TOBA_001/inidata/rays.dat. Picture is from
/PICS/TOBA_001/TRUE_MOD/hor/syn1_ M.png (M=16)

13
Figure 13. Initial model of P anomalies in two vertical
sections which was used to create the real dataset in
TOBA_001/inidata/rays.dat. Picture is from
/PICS/TOBA_001/TRUE_MOD/vert/syn_dv_
M.png (M=1,2)

3.1. MODEL: TOBA_001/model_01

The processing of this model corresponds to inversion of real data. However, the model has been
created artificially using synthetic model shown in Figure 12 (the model parameters can be seen in
TOBA_001/MOD_INI_).

Table 3.1. Parameters of inversion for model TOBA_001/model_01

Data processing real
Inversion scheme Vp-Vs
Number of iterations 5
1D optimization YES
Parameterization nodes
Figure with results 14-15
Approximate calculation time: 1 hour 45 min
Size of the model folder after calculations : 30 Mb

14
Figure 14. Inversion results for P velocity anomalies in 6 horizontal sections based on model
TOBA_001/model_01. Pictures are from:
/PICS/ TOBA_001/model_01/hor/dv15_ M.png, (M=16)

Figure 15. Reconstructed P anomalies in two vertical

sections based on model: TOBA_001/model_01.
Picture is from:
/PICS/TOBA_001/model_01/vert/ver_15 M
.png (M=1,2)

15
4. AREA: TOBA_REA

This is a real dataset uses from LET experiment in the area of Toba Caldera (N.Sumatra). This model
was published in Koulakov et al., (2009, JGI).

Initial dataset:
Number of stations 31
Number of events 505
Number of P rays: 4122
Number of S rays: 2936
Dataset used for inversion:
Number of stations 30
Number of events 390
Number of P rays: 3378
Number of S rays: 2467

4.1. MODEL: TOBA_REA/REAL_VPX

Here we present the results of inversion for P anomalies and Vp/Vs ratio for the area of Toba caldera
based on real data. The resulting values of Vp/Vs ratio are shown in horizontal and vertical sections
in Figures 16 and 17.

Figure 16. Inversion results for P velocity anomalies in 6 horizontal sections based on model
TOBA_REA/real_VPX. Pictures are from:
/PICS/TOBA_VPX/real_VPX/hor/abs_vpvs_5 M.png, (M=16)

16
Table 3.1. Parameters of inversion for model BOLIVAR_/MODEL_01
Data processing REAL DATA
Inversion scheme Vp-Vp/Vs
Number of iterations 5
1D optimization YES
Parameterization nodes
Figure with results 16-17
Approximate calculation time: 2 hours
Size of the model folder after calculations : 28 Mb

Figure 17. Reconstructed values of Vp/Vs ratio

in two vertical sections based on model:
TOBA_REA/real_VPX. Picture is from:
/PICS/TOBA_REA/real_VPX/vert/ab
s_VpVs_5 M.PNG (M=1,2)

17
5. Tests for the GAP criterion, AREAS: GAP__180,
GAP__280, GAP__360

These three tests were produced to check the efficiency of the GAP criterion which is used in most of
LET studies. These examples are constructed based on a real network of stations deployed in Central
Java in the framework of the MERAMEX Project (Koulakov et al., 2007, 2009). In these tests the
data were supplemented with some out-of-network events which were taken from the ISC catalogue.
Maximal epicentral distance to the center of the study area was 5 degrees. We considered three
datasets with different selection criteria:
1. Only events with GAP<180 are taken into consideration.
2. Events with GAP<280 are taken into consideration.
3. All events in the circle of 5 degrees around the network are used

These models are submitted to BSSA in the paper Out-of-network events can be of great use for
Local Earthquake Tomography.

5.1. MODEL: GAP__180/model_01

Here we present the results of inversion with the use of only events with GAP<180.

General information about the data:

(obtained by project /VISUAL/INFO_DATA
Initial dataset:
Table 5.1.ofParameters
Number stations of inversion for model GAP__180/model_01
169
Number of events 130
Data processing real
Number of P rays: 4879
Inversion scheme Vp-Vs
Number of S rays: 2668
Number of iterations 5
Dataset used for inversion:
1D optimization YES
Number of stations 169
Parameterization nodes
Number of events 130
Figure with results 17-18
Number of P rays: 4875
Approximate calculation time: 1 hour
Number of S rays: 2668
Size of the model folder after calculations : 68 Mb

18
Figure 17. Inversion results for P velocity anomalies in 3 horizontal sections based on model
GAP__180/model_01. Pictures are from:
/PICS/GAP__180/model_01/hor/dv15_ M.png, (M=16)

19
Figure 18. Reconstructed P anomalies in two vertical sections based on model:
GAP__180/model_01. Picture is from:
/PICS/GAP__180/model_01/vert/ver_15 M.png (M=1,2)

5.2. MODEL: GAP__280/model_01

Here we present the results of inversion with the use of only events with GAP<280.

General information about the data:

(obtained by project /VISUAL/INFO_DATA
Initial dataset:
Number of stations 169
Number of events 343
Number of P rays: 13060
Number of S rays: 6649
Dataset used for inversion:
Number of stations 169
Number of events 343
Number of P rays: 13055
Number of S rays: 6649

Table 5.2. Parameters of inversion for model GAP__280/model_01

Data processing real
Inversion scheme Vp-Vs
Number of iterations 5
1D optimization YES
Parameterization nodes
Figure with results 19-20
Approximate calculation time: 3 hours 30 min
Size of the model folder after calculations : 94 Mb

20
Figure 19. Inversion results for P velocity anomalies in 3 horizontal sections based on model
GAP__280/model_01. Pictures are from:
/PICS/GAP__180/model_01/hor/dv15_ M.png, (M=1,3)

21
Figure 20. Reconstructed P anomalies in two vertical sections based on model:
GAP__280/model_01. Picture is from:
/PICS/GAP__280/model_01/vert/ver_15 M.png (M=1,2)
5.3. MODEL: GAP__360/model_01

General information about the data:

(obtained by project /VISUAL/INFO_DATA
Initial dataset:
Number of stations 169
Number of events 838
Number of P rays: 31954
Number of S rays: 15797
Dataset used for inversion:
Number of stations 169
Number of events 828
Number of P rays: 31642
Number of S rays: 15606

Table 5.3. Parameters of inversion for model GAP__360/model_01

Data processing real
Inversion scheme Vp-Vs
Number of iterations 5
1D optimization YES
Parameterization nodes
Figure with results 21-22
Approximate calculation time: 11 hours
Size of the model folder after calculations : 161 Mb

22
Figure 21. Inversion results for P and S velocity anomalies in horizontal section based on model
GAP__360. Pictures are from: /PICS/GAP__360/model_01/hor/dv15_ M.png,
(M=1,3).

Figure 22. Reconstructed P anomalies in two vertical sections based on model:

GAP__360/model_01. Picture is from:
/PICS/GAP__360/model_01/vert/ver_15 M.png (M=1,2)
23
7.1. MODEL: TURKEY__/model_01

General information about the data:

Initial dataset:
Number of stations 57
Number of events 2454
Number of P rays: 23265
Number of S rays: 13182
Dataset used for inversion:
Number of stations 40
Number of events 2452
Number of P rays: 23237
Number of S rays: 13175

Table 9.1. Parameters of inversion for model TURKEY__/model_01

Data processing real
Inversion scheme Vp-Vs
Number of iterations 5
1D optimization YES
Parameterization nodes
Figure with results 24
Approximate calculation time: 50 min
Size of the model folder after calculations : 15 Mb

Figure 24. Inversion results for P velocity anomalies in 2 horizontal sections based on model
TURKEY__. Pictures are from: /PICS/TURKEY__/model_01/hor/dv15_ M.png,
(M=1,2).

24
8.1. MODEL: EGYPT_02/BOARD_01

General information about the data:

(obtained by project /VISUAL/INFO_DATA
Initial dataset:
Number of stations 228
Number of events 3170
Number of P rays: 26930
Number of S rays: 5518
Dataset used for inversion:
Number of stations 69
Number of events 1124
Number of P rays: 14199
Number of S rays: 2920

Table 10.1. Parameters of inversion for model EGYPT_02/BOARD_01

Data processing real
Inversion scheme Vp-Vs
Number of iterations 5
1D optimization YES
Parameterization nodes
Figure with results 25

Figure 25. Inversion results for P velocity anomalies in 2 horizontal sections based on model
EGYPT_02. Pictures are from: /PICS/EGYPT_02/BOARD_01/hor/dv15_ M.png,
(M=1,2).

25