Problem Running SH Plane-Wave Simulations

Hi everyone!

I would like to run plane-wave simulations of SH waves in SPECFEM2D. As mentioned in its user manual, I set ‘P_SV’ and ‘initialfield’ parameters in DATA/Par_file to ‘.false.’ and ‘.true.’ respectively. Further, I set ‘source_type’ parameter in DATA/SOURCE to ‘2’ to select the case of S wave. However, when I run the solver, I get weird simulations that do not correspond to SH plane-waves.

Can you please let me know where I am going wrong?

You may note that when I set ‘P_SV’ parameter (in DATA/Par_file) to ‘.true.’ and the remaining parameters same as above, I am able to successfully simulate propagation of SV plane-waves.

Thanks and Kind Regards,
Varun Singla

Research Scientist,
School of Earth Sciences, University College Dublin

could you post your setup files (Par_file, SOURCE) and the output to check? and how does it look when you would use source type 1 for your SH-wave simulations?

Sure. Being a new user, however, I am unable to append these files here and have thus uploaded these on Gdrive:

Source_type_2_files

This is a segment of the output that I get in the terminal:


Preparing initial field for plane wave source
Implementing an analytical initial plane wave…

*** calculation of the initial plane wave ***

To change the initial plane wave, change source_type in DATA/SOURCE and use:
1 or 4 - for a plane P wave
2 or 5 - for a plane SV wave
3 - for a Rayleigh wave
6 - for a mode initial displacement

selected source_type = 2
anglesource = 0.17453292519943295

initial SV wave of 10.000000000000000 degrees introduced.

angle source = 0.17453292519943295

reflected convert plane wave angle: 22.459233779924258
mesh dimensions:
x min/max = 0.00000000 5000.00000
z min/max = 0.00000000 5000.00000

Number of grid points = 160801

You can modify the location of the initial plane wave by changing xs and zs in DATA/SOURCE.
for instance: xs= 2500.0000000000000 zs= 5000.0000000000000 (zs can/should be the height of the free surface)

Max norm of initial elastic displacement = 2.00748444

Preparing source time function
time stepping scheme: Newmark
time stepping stages: 1
time step size : 1.00000005E-03

number of time steps: 3200
initital field : T

Preparing image coloring
allocating color image arrays
locating all the pixels of color images
total number of image pixels = 648830
done locating all the pixels of color images
coloring image background based on vp

done, preparation successful


The medium is homogeneous, elastic, and isotropic with width=height=8 km. Simulation is for right-travelling waves incident at an angle of 10 degrees. Ricker wavelet of frequency 6 Hz is chosen as source function. It appears from the above output that instead of incident SH plane-waves, SV plane-waves have been considered by SPECFEM2D. This is weird considering that I have set P_SV=’.false.’ in Par_file.
Below is a contour image of the displacements at an initial time-step, which also indicates that the case of an SH plane-wave has not been considered (there should be only 2 plane-waves: one incident and one reflected, whereas this image shows three).

It looks similar for source type 1, i.e., instead of incident plane SH waves, incident plane P waves are considered by SPECFEM2D. Weirdly enough, the output files correspond to displacement along only y-direction for both source type 1 and 2 (displacements for incident P and SV plane-waves should occur along x- and z-directions only).

Much thanks for taking the time to go through my files!
Varun