Source Parameters for the Community Stress Drop Validation Study Derived from Stopping Phases

The community-led stress drop validation study initiated in early 2021 aims at understanding and resolving differences in stress drop measurements using a consistent dataset of the 2019 Ridgecrest earthquake sequence. Stress drop values are commonly estimated by fitting theoretical source models to direct phase (P- or S-wave) or coda wave spectra. In this work, we contribute to the community-led study by using an alternative approach to estimate fault dimension and rupture velocity (the main parameters modulating stress drop) that employs an inversion method based on stopping phases. In estimating source dimension and rupture velocity from stopping phases, we assume an elliptical or circular rupture from which two high-frequency stopping phases radiate. The approach measures the differential time between the two observed stopping phases following a direct P- or S-wave arrival at multiple stations. The rupture velocity and fault dimension are then constrained by an inversion based on observations at all stations with observable differential times. The main benefit of this inversion method lies in its independence from corner frequency estimates that can be strongly affected by bandwidth limitations and travel-path effects. The method has been shown to resolve source properties of small earthquakes (M < 3.0) in specific cases where instrumental coverage and data quality permit. The Ridgecrest dataset offers a unique opportunity to explore the applicability, potential benefits, and limitations of the method due to a wide range of earthquake magnitudes, variety of seismic instruments, and extensive body of supporting research related to the sequence. We use state-of-the-art optimization techniques and high-performance computing to tackle known problems of the inversion, such as ambiguities in detecting stopping phases and the high number of unknown variables. First results show clear detection of stopping phases for a subset of earthquakes that indicate stress drop values ranging between 0.1 - 10 MPa. We will also present stress drop estimates using frequency-based methods, such as single spectra and spectral-ratio fitting, to provide comparative estimates using different approaches.