Risk Assessment of Building Structural Vulnerability in California Based on Non-Ergodic and Ergodic Probabilistic Seismic Hazard Analysis

Session: Forthcoming Updates of the USGS NSHMs: Hawaii, Conterminous U.S. and Alaska [Poster]
Type: Poster
Date: 4/28/2020
Time: 08:00 AM
Room: Ballroom
Description: 

Assessing the seismic risk of buildings is of great benefit for the development of contingency strategies before the earthquakes. Traditional approaches for building risk assessment utilize ergodic ground motion models (GMMs) developed for broad regions such as GMMs in NGA-West2 to estimate ground motion intensity measures (IMs) and combine these estimates with fragility functions to compute the damage state probability of a specific building type. However, the influence of the ergodic assumption of traditional GMMs on the estimated risk has not been explored. With the growth of ground motion database, significant regional differences (on the scale of 10s of km) for recorded ground motions are observed, which does not support an ergodic assumption. These spatial difference in ground motions can be addressed by using a non-ergodic GMM with location-specific coefficients, which is also known as varying coefficient model (VCM) (Abrahamson et al., 2019). It has been proved that non-ergodic GMMs can achieve a significant reduction in aleatory variability of ground motion prediction, and there is a move from ergodic GMMs to non-ergodic GMMs in probabilistic seismic hazard analysis (PSHA). It is, however, unknown the influence of the reduction of aleatory variability on risk estimates. This study explores the effect of using non-ergodic GMMs on the damage analysis of buildings as compared to the use of traditional GMMs. Ground motion IMs and the associated uncertainties are estimated for 3 areas of different seismicity (varying from low to moderately large) in California (San Jose, San Luis Obispo and North East California) using: 1) a non-ergodic PSHA and 2) a traditional ergodic PSHA. Then these IMs are taken as input for the fragility functions of different building types, as defined in HAZUS manual. The estimated IMs and fragility functions are used to perform risk estimates considering both the non-ergodic and ergodic scenarios. Finally, insights in terms of potential effects of the non-ergodic assumptions in areas of different seismicity on risk estimates are shared.

Presenting Author: Jorge Macedo

Authors