Comparisons Between Single and Weighted Average NGA-West2 GMPEs in ShakeMap

Date: 4/25/2019

Time: 06:00 PM

Room: Grand Ballroom

ShakeMap uses both observed input data and estimates of ground motions from ground motion prediction equations (GMPEs) to represent the ground shaking experienced in an earthquake. In the California Integrated Seismic Network, Next Generation Attenuation (NGA) GMPEs developed for the Western US are used to estimate ground motions in ShakeMaps. While NGA-West2 GMPEs are generally consistent with each other, the choice of an appropriate single GMPE in ShakeMap among the NGA-West2 GMPEs may depend on the ground motion prediction model parameters available immediately after an earthquake. In the initial post-quake interval, some model parameters needed in the GMPEs, such as hypocentral depth, rupture extent or fault orientation may not yet be available. In the absence of these parameters, defaults or proxies may be used, yet it is not clear in a ShakeMap context which GMPEs perform best in these situations. Furthermore, recent developments in ShakeMap version 4.0 allow the use of either a single GMPE or a weighted average of two or more GMPEs. This study investigates the behavior of four of the five NGA-West2 GMPEs and compares them with estimates obtained from an equally-weighted multi-GMPE approach for different intensity measure types. By evaluating their sensitivities to perturbations of the model parameters, the study will also determine if the multi-GMPE gives a good estimate of ground motions. This study will further investigate the variability of the multi-GMPE ground motion estimates with different region-specific site conditions. Adjustments to the weighting of the four NGA-West2 GMPEs are made by comparing GMPE ground motion-based estimates with recorded data from California earthquakes for a range of magnitudes.

Presenting Author: Lijam Hagos

Authors