SW4 Mesh Development

From SCECpedia
Jump to navigationJump to search

Goal: Create NorCal CyberShake Mesh for SW4

Mesh Region

Region should match the SFBR model https://www.usgs.gov/programs/earthquake-hazards/science/3-d-geologic-and-seismic-velocity-models-san-francisco-bay

  • ne_coord (-122.562365, 39.174505)
  • nw_coord (-123.858493, 38.424179)
  • sw_coord (-121.922036, 36.320331)
  • se_coord (-120.644051, 37.050062)
  • 140km x 290km x 50km

Config Script

  • input 4 corners (and projection) and spacing
  • calc rotation angle
  • distance between points
  • calculate nx, ny, nz based on spacing

Define Region

  • origin
  • rotation angle
  • distance x
  • distance y
  • distance z
  • spacing

Low Res UCVM2Mesh Configuration for SFBR Model

To test our mesh formats for SW4, we will test with a low resolution

  1. rot=-53.9
  2. x0=-123.858493
  3. y0=38.424179
  4. z0=0.0
  5. spacing=500.0
  6. Dist x = 140,000
  7. Dist y = 290,000
  8. Dist z = 40,000
  9. nx=280
  10. ny=580
  11. nz=80

Full Res UCVM2Mesh Configuration for SFBR Model

  1. rot=-53.9
  2. x0=-123.858493
  3. y0=38.424179
  4. z0=0.0
  5. spacing=25.0
  6. Dist x = 140,000
  7. Dist y = 290,000
  8. Dist z = 60,000
  9. nx=
  10. ny=
  11. nz=

SF CVM

These are the coordinates of the detailed USGS SFBA VM:

ne_coord (-122.562365, 39.174505) nw_coord (-123.858493, 38.424179) sw_coord (-121.922036, 36.320331) se_coord (-120.644051, 37.050062)


Regarding the azimuth of the velocity model in the SFBA, we use a rotation of 143.638º, assuming an initial rectangle aligned with the geographical coordinates and anchor at the south-west corner. After applying this rotation, this point at the origin becomes the north-east corner of the domain.
can you discretize the mesh in 100 m in horizontal and 25 m in vertical, following a cube? Eventually at larger depths the discretization is coarser which will allow to same mamory, but I do not know yet how the model is defined in depth (I think at a some points the blocks are defined every 200 m in depth, but I haven't explored this feature of the velocity model).

Example from CyberShake

We don't actually generate a ucvm2mesh config file as part of the process of creating CyberShake meshes, but I've created what I think is the correct ucvm2mesh file for USC. The rotation angle is -55 degrees, and the 4 corners should be about:

   -122.46716     34.52173 (x=      0.000 y=      0.000)
   -119.66138     37.97316 (x=    460.000 y=      0.000)
   -113.59525     34.44267 (x=    460.000 y=    672.000)
   -116.52591     31.13175 (x=      0.000 y=    672.000)

Sim Volume CyberShake Site s3446

  • The yellow box, the volume for site s3446 - has the following specs:
  • Dimensions: 460.8 x 774.4 x 50.56 km
  • Rotation angle: 36 degrees counterclockwise
  • Corner points:
    •       -126.18649, 39.75063 (W)
    •       -121.85281, 42.27791 (N)
    •       -116.72395, 36.56162 (E)
    •       -120.90294, 34.22243 (S)

Related Entries