← Back to Sessions

Spatial Correlation of Earthquake Ground Motion from 3D Physics-Based Ground Shaking Scenarios

Session: Physics-Based Earthquake Rupture Modeling and Strong Motion Simulations II

Type: Oral

Date: 4/23/2021

Presentation Time: 02:30 PM Pacific

Description: 

One of the key issues in the seismic risk assessment of spatially distributed portfolios or infrastructural systems in large urban areas is the modeling of the spatial correlation of ground motion intensity measures. Recently, several spatial correlation models have been proposed on empirical basis using wide datasets of earthquake recordings in different areas worldwide and, therefore, they may be poorly representative of specific regions and/or near-source effects. Three-dimensional physics-based numerical simulations (PBS) of seismic ground motion allow to investigate region-specific features of spatial correlation, with a level of detail which could not be possible using sparse recordings, and to explore the dependence on physical factors in near-fault conditions, such as source directivity, path and site effects. This contribution focuses on the analysis of spatial correlation of earthquake ground motion from a wide set of broadband PBSs, generated through the spectral element code SPEED, coupled with an Artificial Neural Network technique for the high-frequency part. Geostatistical tools, based on the computation of both the semi-variogram and the Pearson’s correlation map, are employed to assess the correlation between spatially distributed response spectral accelerations (SA) at different vibration periods. The numerical dataset encompasses seven areas worldwide (Po Plain, L’Aquila, Marsica, Sulmona, Norcia, in Italy; Istanbul, in Turkey; Thessaloniki, in Greece) and moderate-to-severe earthquake scenarios (M=6-7+) in near-source conditions. The results point out that spatial correlation may be strongly region- and scenario-specific. Besides, factors, such as magnitude, source directivity, path propagation, directionality (fault normal versus fault parallel), may affect significantly the spatial correlation, making critical the use of isotropic and stationary models especially in near-fault conditions.

Key words: spatial correlation; 3D physics-based numerical simulations; geostatistical analysis; near-fault conditions

Presenting Author: Jiayue Lin

Student Presenter: Yes

Authors