McVerry and Zhao et al have problematic code to find middle of rupture surface

[Kevin Milner]

McVerry? and Zhao et al currently calculate the center location of a surface by averaging the lat/lon/depth values of all points on the surface. This would work OK for planar faults, but this center point would be off the fault surface for faults with any curvature.

It appears that this center location is used as the hypocenter in the GMPEs.

[Peter Powers]

This can probably be closed as won't fix. Hypocentral depths are are used in a variety of attenuation relations independent of lat-lon position; the existing averaging algorithms are likely fine assuming they're restricted to the attenuation relationships mentioned.