Progress in measuring spatial correlations in ground motion intensity

Jack W. Baker, Maryia Markhvida, & Yilin Chen

Submitted September 13, 2020, SCEC Contribution #10023

For nearly 20 years, studies of spatial correlations in ground motion intensity have been an area of active research. Models of spatial correlation are integral for assessing seismic risk to distributed infrastructure systems, portfolios of insured properties, or any other systems affected by shaking at more than one location. These spatial correlation models have been calibrated using ground motion data from Japan, USA, Italy, New Zealand, and elsewhere around the world. Models have been developed for a range of ground motion intensity metrics, including spectral acceleration at a number of periods, peak ground acceleration, peak ground velocity, and Arias Intensity. And studies have been performed to evaluate whether these correlations depend upon the tectonic region, the underlying geotechnical conditions, or the properties of the causal earthquake rupture.

Despite the importance of these models, their development is significantly constrained by the availability of suitable recorded ground motions at closely spaced stations. With that context, this paper reviews recent developments by the authors in two areas: (1) new statistical techniques for estimating spatial correlations from data, (2) measurement of spatial correlations from physics-based ground motion simulations. These developments will be described, their potential impact on seismic risk assessment will be explained, and prospects for future developments will be discussed.

Key Words
ground motion; spatial correlation; regional risk assessment

Baker, J. W., Markhvida, M., & Chen, Y. (2020, 09). Progress in measuring spatial correlations in ground motion intensity. Oral Presentation at 17th World Conference on Earthquake Engineering, 17WCEE.