UCERF3 users Gabriele Coccia and Paolo Bazzurro discovered an issue with subsection ordering in the electronic subplements to the UCERF3-TD model. This also affected a few of the figures from the final BSSA paper.
Sub section probability CSV data files, posted in the UCERF3-TD electronic supplements (for example: http://opensha.usc.edu/ftp/UCERF3-TimeDependentSupplements/all_5yr/sub_section_probabilities.csv), were originally sorted by name. This led to an incorrect ordering of subsections on faults with 10 or more subsections, for example:
Fault x Subsection 1 Fault x Subsection 10 Fault x Subsection 11 Fault x Subsection 2 Fault x Subsection 3 ... Fault x Subsection 9
The data, however, was sorted by subsection number (1, 2, 3, etc). All files have been updated, with the original (with the bug) data preserved here. The corrected files were moved into place to replace the original (with bug) files on 4/12/2017.
The following figures from the UCERF3-TD BSSA figures were sensitive to this bug and have been updated online:
a) Mean Time Dep Probs:
b) UCERF3-TD Prob Gain
Figure 13 (deformation model branch sensitivity):
Figure 14 (scaling relationship sensitivity)
Figure 15 (probability model sensitivity)
Copyright ©2021 WGCEP.org and the University of Southern California. All rights reserved.