Getting all traveltimes from 3D rectangular grid

[SquirrelKnight]

Hello,

I am currently working on deriving travel time grids for several different stations for a heterogenous velocity model. I am using the rgrid method to make a 3D grid. From there, I would like to get traveltimes for all nodes on the grid.

So far, I have attempted setting the source as the station, and the receiver as a random point in space within the grid. I get a result that makes sense using grid.raytrace. After doing this, I have attempted grid.get_traveltimes(), however the traveltimes at the same nodes do not match the output of the raytrace. For example, the location of the source should have a traveltime of 0.0 s, but it doesn't. Instead the 0.0s traveltime is located somewhere else entirely in the grid. I made sure to check that I was converting between the x, y, and z coordinates and the actual node points correctly, so that isn't the issue.

This initial test was performed using the FSM method. I was wondering if the only accurate way to get traveltimes to each node point is to actually set each node as a receiver. From what I understand in the documentation, it seems like the traveltimes should be propagated across the grid when first running the raytrace function, am I wrong in thinking so? Your advice is greatly appreciated!

Here is an example of my code, which uses a NonLinLoc velocity grid as the input:

import numpy as np
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
from nllgrid import NLLGrid
import ttcrpy.rgrid as rg

nll_file = '/Volumes/SeaJade 2 Backup/ONC/endeavour_analysis/data/nll/model/END.P.mod.buf'
p_grd = NLLGrid(nll_file)

V = 1 / (p_grd.array / 0.2)

# Get number of grids in x, y, and z directons
nx, ny, nz = V.shape

# Get starting point in x, y, and z directions
x_orig, y_orig, z_orig = p_grd.x_orig, p_grd.y_orig, p_grd.z_orig

# Get spacing in x, y, and z directions
dx, dy, dz = p_grd.dx, p_grd.dy, p_grd.dz

x = np.arange(0,nx*dx,dx) + x_orig
y = np.arange(0,ny*dy,dy) + y_orig
z = np.arange(0,nz*dz,dz) + z_orig

grid = rg.Grid3d(x, y, z, cell_slowness=False, method='SPM')

src = np.array([[-1.0,0.0,0.0]])
rcv = np.array([[1.0,1.0,1.0]])

slowness = 1./V
tt, rays = grid.raytrace(src, rcv, slowness, return_rays=True)

grid_tt = np.array(grid.get_grid_traveltimes())