Inputs

In JSWAP, all of the input parameters are assigned to a struct input2. The following input parameters are needed to run the simulation. Please go through one by one to check the input is correct.

1. dt

Explanation: Time interval.

Type: Float64.

Dimension: [].

Example:

input2.dt=10.0^-3;

2. dx, dy, dz

Explanation: Grid spacing in the x, y and z directions.

Type: Float64.

Dimension: [].

Example:

input2.dx=10.0;
input2.dy=10.0;
input2.dz=10.0;

3. nt

Explanation: Time interval.

Type: Float64.

Dimension: [].

Example:

input2.nt=10.0^3;

4. nx, ny, nz

Explanation: Grid points in the x, y and z directions.

Type: Int32.

Dimension: [].

input2.nx=80;
input2.ny=80;
input2.nz=90;

5. X,Y,Z

Explanation: 3D true coordinate of the model.

Type: Float or Int.

Dimension: [nx, ny, nz]

Example:

input2.Y,input2.X,input2.Z=JSWAP.meshgrid(1:80,1:80,1:90);

6. Stiffness

Explanation: Lame constants, λ and &mu or stiffness matrix (Cij). Either of them is required for input.

Type: Float64.

Dimension: [nx, ny, nz].

Example:

# Lame constants for isotropic media
input2.lambda=ones(80,80,90)*10^9*1.0;
input2.mu=ones(80,80,90)*10^9*1.0;

# Stiffness matrix for anisotropic media
input2.C11=ones(80,80,90)*10^9*3.0;
input2.C12=ones(80,80,90)*10^9*1.0;
input2.C13=ones(80,80,90)*10^9*1.0;
input2.C14=ones(80,80,90)*10^9*0;
input2.C15=ones(80,80,90)*10^9*0;
input2.C16=ones(80,80,90)*10^9*.5;
input2.C22=ones(80,80,90)*10^9*3.0;
input2.C23=ones(80,80,90)*10^9*1.0;
input2.C24=ones(80,80,90)*10^9*0;
input2.C25=ones(80,80,90)*10^9*0;
input2.C26=ones(80,80,90)*10^9*.5;
input2.C33=ones(80,80,90)*10^9*3.0;
input2.C34=ones(80,80,90)*10^9*0;
input2.C35=ones(80,80,90)*10^9*0;
input2.C36=ones(80,80,90)*10^9*0;
input2.C44=ones(80,80,90)*10^9*1.0;
input2.C45=ones(80,80,90)*10^9*0;
input2.C46=ones(80,80,90)*10^9*0;
input2.C55=ones(80,80,90)*10^9*1.0;
input2.C56=ones(80,80,90)*10^9*0;
input2.C66=ones(80,80,90)*10^9*1.0;

7. rho

Explanation: Density.

Type: Float64.

Dimension: [nx, ny, nz].

Example:

input2.rho=ones(80,80,90)*1000.0;

8. inv_Qa

Explanation: Apparent attenuation.

Type: Float64.

Dimension: [nx, ny, nz].

Example:

input2.inv_Qa=ones(80,80,90)*0.0;

8. r1, r2, r3

Explanation: Grid locations of receivers in the x, y and z directions.

Type: Int32.

Dimension: Each of them is a 2-dimensional matrix, with the first direction to be 1 and the second direction as the number of receiver numbers.

Here is an example of 50 receivers.

input2.r1=zeros(Int32,1,50);
input2.r1[:]=30:79;

input2.r2=zeros(Int32,1,50);
input2.r2[:]=30:79;

input2.r3=zeros(Int32,1,50);
input2.r3 .=15;

9. s1, s2, s3

Explanation: Grid locations of sources in the x, y and z directions.

Type: Int32.

Dimension: Each of them is a 2-dimensional matrix, with the first direction to be 1 and the second direction as the number of receiver numbers.

Here is an example of 2 sources.

input2.s1=zeros(Int32,1,2);
input2.s1[:] =[60 60];

input2.s2=zeros(Int32,1,2);
input2.s2[:] =[49 51];

input2.s3=zeros(Int32,1,2);
input2.s3[:] =[49 51];

10. Source signals

Explanation: Source signal in each component. The source can be either directional source, P-wave source or moment-tensor source. There are 2 options:

If one chooses directional source and P-wave source, the input is supposed to be input2.src1, input2.src2, input2.src3 and input2.srcp, corresponding to the source signals in the x, y, z directions and the P-wave components.
If moment tensor source is wanted, M11, M22, M33, M23, M13 and M12 are desired.

Type: Float64.

Dimensions: [nt, number of sources].

11. r1t, r2t, r3t

Explanation: Receiver true locations in the x, y and z directions.

Type: Int32 or Float64.

Dimension: [1, number of receivers].

Example of computation true locations based on above information.

input2.r1t=input2.r1*input2.dx;
input2.r2t=input2.r2*input2.dy;
input2.r3t=input2.r3*input2.dz;

12. Rm

Explanation: One can mute some receivers with this option. 0 - mute, 1 - unmute.

Type: Int32 or Float64.

Dimension: [number of receivers, component].

Component: 1 - x component, 2 - y component, 3 - z component, 4 - pressure component.

If one does not want to mute any receivers, then

input2.Rm=ones(length(input2.r3),4);

13. s1t, s2t, s3t

Explanation: Source true locations in the x, y and z directions.

Type: Int32 or Float64.

Dimension: [1, number of sources].

Here is an example of 50 receivers.

input2.s1t=input2.s1*input2.dx;
input2.s2t=input2.s2*input2.dy;
input2.s3t=input2.s3*input2.dz;

14. lp

Explanation: PML layers.

Type: Int32.

Dimension: [].

For a 10 layer PML,

input2.lp=10;

15. nPML

Explanation: power of PML, normally 2.

Type: Int32.

Dimension: [].

Example:

input2.nPML=2;

16. Rc

Explanation: PML theoretical reflection coefficient.

Type: Float64.

Dimension: [].

Some common coefficients:

PML layers (lp)	Rc
10	10.0^-1
20	10.0^-2
30	10.0^-3
40	10.0^-4

Example:

input2.Rc=.1;

17. PML_active

Explanation: One can set if PML at one edge is working. Each element of the 1 by 6 matrix means if the PML is activated on the edge in an order of xminus, xplus, yminus, yplus, zminus, zplus. "0" is deactivated and "1" is activated.

Type: Float64.

Dimension: [1, 6].

Example for only zminus PML is deactivated:

input2.PML_active=[1 1 1 1 0 1];

18. path

Explanation: Path of master branch on where to store the wavefield vtk and wavefield mat file. This must be a parent path of pathpic, pathmodel and pathwavefield, pathrec.

Type: String.

Dimension: NA.

input2.path=p3;

19. path_pic

Explanation: Path to store wavefield snapshot vtk file. If one assign it "nothing", then no wavefield snapshot will be saved.

Type: String.

Dimension: NA.

input2.path_pic=string(input2.path,"/pic");
input2.path_pic=nothing;

20. path_model

Explanation: Path to store the model, including material parameters and source and receiver locations.

Type: String.

Dimension: NA.

input2.path_model=string(input2.path,"/model");

21. path_wavefield

Explanation: Path to store mat wavefield. No wavefield matfile will be stored if path_wavefield is "nothing".

Type: String.

Dimension: NA.

input2.path_wavefield=string(input2.path,"/wavefield");

22. path_rec

Explanation: Path to store recordings.

Type: String.

Dimension: NA.

input2.path_rec=string(input2.path,"/rec");

23. plot_interval

Explanation: Plot frequency/interval of vtk file. "0" for saving nothing.

Type: Int32.

Dimension: [].

input2.plot_interval=100;

24. wavefield_interval

Explanation: How frequent the wavefield is saved. "0" for saving nothing.

Type: Int32.

Dimension: [].

input2.wavefield_interval=0;

An example of input file is given here.