Review NGA-SUbduction zTor, zHyp usage

added 2023 GMM labels

mentioned in issue #86 (closed)

changed milestone to %Beta release of CONUS NSHM

changed milestone to %Beta release of CONUS NSHM 2

Equations relating zTor to zHyp for Parker et al. (2021) are required

For PSBAH21, the current implementation incorrectly passes through zTor in place of zHyp. Slab sources in nshmp-haz are modeled as gridded point sources with depths to the top of the slab from SLAB2 (we take the top of the slab as zTor). To properly use the zTor --> zHyp_bar conversion provided in Parker et al. (2021), we need to make decisions about the finite fault representation that we might use. This in turn raises questions about the dip of the subducting slab and the dip of possible conjugate fault pairs; these are modeling decisions that the PEER PSHA code verification project (Hale et al., 2018) discussed at length, and there were a dizzying array of implementation variations. To keep things simple we are going to assume a vertical fault at each grid point and a slab with vertical thickness of 6.5 km (zTor + 0.48*6.5). It should be noted that the Cascadia data points in Parker et al. (2021) are shallower relative to zTor (fig 7b) than average for the dataset that is dominated by Japan data; does this reflect thinner young crust being subducted? perhaps). At this point we are not going to apply additional epistemic uncertainty to zHyp as it requires running multiple scenarios.

Resolved in !381 (merged)

closed