- Sung, James, Jason
Aims
- Wairau and WaipukPouk at 100m
- Finish Wairau 200m IM calc
- Try 0km, 10km smoothing distance to see the effect of blending/smoothing
Added during Sprint
- Support intended varying& high depth resolution
Highlights
...
elev=-9000 → surf_tomography_vp_elev-9.in
elev=-8750 → surf_tomography_vp_elev-8.in
elev=-8500 → surf_tomography_vp_elev-8.in
elev=-8250 → surf_tomography_vp_elev-8.in
elev=-8000 → surf_tomography_vp_elev-8.in
Shallow depth has 250m granularity, while NZVM is limited to 1km resolution. -8750, -8500, -8250 data are overwritten.
Question remains: Does it matter? Robin thinks not.
Completed
- Fix Chow tomography integration code to produce surf_tomography_vp_elev-8.in, surf_tomography_vp_elev-8p25.in, surf_tomography_vp_elev-8p50.in, surf_tomography_vp_elev-8p75.in
- Fix NZVM to process depths in floating point number and increase max tomo surfaces
Issues
NZVM calculation runs into malloc error (seg fault) if run interactively → Should be submitted as a job
EMOD3D output is filled with NaN values ← Found Mom(src) becomes infinity at some procs, which gets propagated/amplified everywhere via MPI collective calls (Details)
Outstanding
Initial Plan for next sprint
Fix wrong total moment tensor