A Geodesy- and Seismicity-Based Local Earthquake Likelihood Model for Central Los Angeles

Date: 4/24/2019

Time: 04:15 PM

Room: Pike

We estimate long-term-average earthquake likelihoods by magnitude in central Los Angeles using a recently developed model of interseismic strain accumulation and the 1932-2017 SCEDC seismic catalog. We assume that on the long-term average, 1) mainshocks obey a log-linear magnitude-frequency distribution (MFD) up to a maximum magnitude and follow a Poisson recurrence process, 2) aftershocks obey a log-linear MFD and “Bath’s law” (largest aftershock 1.2 magnitude units below the mainshock), and 3) these mainshocks and aftershocks (and aseismic slip) collectively release seismic moment at a rate balancing the geodetic interseismic loading. We develop a method to use these constraints and the seismic catalog to probabilistically estimate the long-term MFD of earthquakes. The method generates a comprehensive suite of long-term MFDs obeying these assumptions, assesses how likely a long-term system obeying each MFD would be to produce earthquakes with an MFD like that of the 86-year SCEDC catalog over an 86-year period, and use these likelihoods to build the probability density functions of M_max, the b-value of the MFD, and earthquake rates at any magnitude. We estimate M_max=6.8 +1.05/-0.4 (every ~300 years) or M_max=7.05 +0.95/-0.4 assuming a truncated or tapered Gutenberg-Richter MFD, respectively. Our results imply that, for example, the (median) likelihood of a Mw≥6.5 mainshock in central LA is 0.2% in one year, 2.0% in 10 years, and 18% in 100 years. The effective "non-declustered" b-value of the best long-term models is 0.9-1. The actual parameter bin this method is essentially the "declustered" b-value as it governs the MFD of mainshocks; it is close to published estimates for the declustered b-value in California (~0.8-0.9). We find that if this b is instead modeled as governing a single MFD for all earthquakes (no mainshock-aftershock distinction) as the "non-declustered" b-value, M_max changes by only 0.1 and the long-term model is virtually unchanged.

Presenting Author: Chris Rollins

Authors