Opened 10 years ago
Closed 10 years ago
#446 closed defect (fixed)
Finish QuadSurface implementation and tests
| Reported by: | Kevin Milner | Owned by: | |
|---|---|---|---|
| Priority: | major | Milestone: | OpenSHA 1.4 |
| Component: | sha | Version: | |
| Keywords: | Cc: |
Description
QuadSurface? has been mostly finished and tested but final test thresholds need to be set for comparisons against evenly gridded representations. Also, Peter thinks the DistanceX calculation could be more efficient
Quoting Peter on 12/10/13
Kevin, I looked into you're QuadSurface? implementation of rX. I think both the current gridded implementation and yours are overkill. rX can be computed completely independently of quads and rotations. Assuming the right-hand-rule is adhered to in fault and dip definitions, then I think the code below would work.
Attachments (1)
Change History (2)
Changed 10 years ago by
| Attachment: | dist_x_compare.png added |
|---|
comment:1 Changed 10 years ago by
| Resolution: | → fixed |
|---|---|
| Status: | new → closed |
I think we're good to go. One important note for distance X calculations is that average strike is used for the infinite fault extension (just like the gridded surfaces), but that extension is done in cartesian coordinates rather than great circle (which is the case for gridded). This causes slight differences in sign/values directly off the end of faults. See attached figure with quad distances on top, gridded on bottom.
distance x conmparison, quad surfaces on top, gridded on bottom