Combining Dynamic Rupture Simulations with Ground-Motion Data to Characterize Seismic Hazard from M-w 3 to 5.8 Earthquakes in Oklahoma and Kansas

摘要

Many seismically active areas suffer from a lack of near-source ground-motion recordings, making ground-motion prediction difficult at distances within similar to 40 km from an earthquake. We aim to aid the development of near-source ground-motion prediction equations (GMPEs) by generating synthetic ground-motion data via simulation. Building on previous work using point-source moment tensor sources to simulate small (M-w 3-4) earthquakes for a target region encompassing north central Oklahoma and south central Kansas, we perform dynamic rupture simulations of earthquakes up to M-w 5.8. We introduce complexity into the rupture process by adding stochastically generated heterogeneity to initial stress conditions. Our simulated ground-motion data are added to a catalog of recorded ground-motion data to construct a composite recorded-simulated ground-motion catalog that we use to develop a GMPE for the target region. This procedure can be generalized and used for GMPE development in other regions with near-source ground-motion data scarcity, which could directly benefit critical applications such as the National Seismic Hazard Maps produced by the U.S. Geological Survey (USGS).