Impact of Declustering on Probabilistic Seismic Hazard Estimates in the United States
Session: Probabilistic Seismic Hazard Assessment: Where Do We Go from Here? II
Type: Oral
Date: 4/20/2021
Presentation Time: 03:30 PM Pacific
Description:
Hazard estimates from background seismicity differ by more than a factor of two, depending on the way that the catalog is declustered. In this study, we begin with the complete seismicity catalogs used by the U. S. Geological Survey (USGS) to develop the background seismicity model for the 2014 National Seismic Hazard Model. The first catalog that we use is declustered by the USGS based on the approach and parameters suggested by Gardner and Knopoff (1974). We also consider declustered catalogs using the nearest-neighbor (NN) approach of Zaliapin and Ben Zion (2020). The NN approach randomly thins the catalog to obtain a stationary background earthquake rate, while preserving its spatial variability. The size of a declustered catalog is controlled by the estimated probability that a given event is independent, and parameterized with a threshold a0. Smaller a0 (e.g. a0 = –1) identifies as background only events that are highly likely to be independent, while larger a0 (e.g. a0 = 1) includes more events that have a higher probability of being part of a cluster. The number of background events and the hazard estimates thus increase with a0. Our results suggest that the hazard is practically independent of the variation between different realizations of the declustered catalog for the same a0. At a sample location in Nevada, where the background dominates the hazard, the hazard curve from declustering with a0 = 1 is similar to the hazard (PGA at 2% in 50 years) obtained by the USGS, and decrease to about half of that value when the threshold a0 decreases to –1.
Gardner, J. K. & Knopoff, L. (1974). Is the sequence of earthquakes in southern California, with aftershocks removed, Poissonian? Bulletin of the Seismological Society of America, 64, 1363–1367.
Zaliapin, I., & Ben‐Zion, Y. (2020). Earthquake declustering using the nearest‐neighbor approach in spacetime‐magnitude domain. Journal of Geophysical Research: Solid Earth, 125, e2018JB017120.
Presenting Author: Emily L. Maher
Student Presenter: Yes
Authors
Emily Maher Presenting Author Corresponding Author emaher@nevada.unr.edu University of Nevada, Reno |
Ilya Zaliapin zal@unr.edu University of Nevada, Reno |
John Anderson jga@seismo.unr.edu University of Nevada, Reno |
|
|
|
|
|
|
Impact of Declustering on Probabilistic Seismic Hazard Estimates in the United States
Category
Probabilistic Seismic Hazard Assessment: Where Do We Go from Here?