NonLinLoc Transformation Projection Question

[ev-anderson]

Hi Anthony,

I have a 3d velocity model for my study region about 80 x 115 x 50 km. I generate my model using a bathymetry grid in UTM coordinates, projecting a 1d velocity model at each bathymetric grid node, and interpolating in between each node. I then use nllgrid to write this into a NLLOC model file. When running the locations, I've been using TRANS MERC as my transformation. Would the use of UTM initially for the velocity model construction and the selection of TRANS MERC in the control file have an affect on my locations?

Evan

[alomax]

I think that UTM and TRANS MERC can be the same, if the correct, corresponding parameters are used for each. See https://docs.generic-mapping-tools.org/6.0/cookbook/map_projections.html#transverse-mercator-projection-jt-jt

Is there some way you can compare numerically or graphically the original model in km or meters xyz on the bathymetric grid node with that written as a NLLOC model file? Grid2GMT can help with the NLL model.
If there is some distortion error (in distance), this error divided by the local velocity would give an approximate travel-time error. If this time error multiplied by, e.g., the typical number of cells between the stations and source is much less than the pick uncertainty, then there may be no loss of precision. But this is all a very rough estimation!

Anthony