Extended Stochastic Finite-Fault Ground-Motion Simulation Incorporating Variable Stress Parameter and Stochastic Slip Models

The stochastic finite-fault (SFF) modeling approach is widely used to simulate earthquake ground motions. In its original formulation, the stress parameter, one of the key source characterizations, is assumed to be spatially uniform across the fault plane. However, empirical source-inversion and theoretical studies indicate that stress parameter varies spatially across the fault plane for a given earthquake. To address this limitation, the SFF method is extended to incorporate spatially variable stress parameter derived from earthquake slip models. In the extended framework, a more realistic representation of rupture complexity is achieved by generating spatially correlated synthetic slip models using the von Kármán autocorrelation function and by converting them into spatially variable stress parameter distributions. The 2014 South Napa earthquake (moment magnitude 6.07) is selected for method development and evaluation because of the availability of dense strong-motion recordings in the near-fault region. Model performance is assessed through residual analyses between observed and simulated ground motions. Sensitivity analyses indicate that the extended SFF method is particularly responsive to slip-model variability at short periods of response spectra in the near-fault region. The results further demonstrate that the proposed approach models the spatial dependence of simulated ground motions associated with asperity locations, leading to a more physically consistent representation of source effects in SFF simulations.