Exploring the Dependence of Corner Frequency Estimates on EGF-selection Criteria using the 2019 Ridgecrest Earthquake Sequence

Abstract:

Stress drop is a commonly used measure of source complexity and an important source parameter for modeling and understanding strong ground motion, and hence, is fundamental to understanding the rupture process and calculating seismic hazard. In this study, we use an Empirical Green’s Function (EGF) method to estimate source parameters associated with the 2019 Mw6.4 and Mw7.1 Ridgecrest Earthquake Sequence in the Southern Walker Lane, California, USA. The traditional EGF approach uses highly similar, smaller magnitude, collocated earthquakes as Green functions to isolate the source spectra of a target event as the spectral ratio between two events. The corner frequency can be measured from the ratio and assuming a simple circular source model (e.g., Brune) and a rupture velocity, source dimension and stress drop can be calculated. Although the estimation of a stress drop from the corner frequency is straightforward, under these assumptions, it is harder to be certain that the corner frequency measurements are reliable and accurate. Furthermore, the stress drop is proportional to the cube of the corner frequency and the measurement uncertainties with corner frequency are tripled when calculating stress drop. Because the selection and quality of potential EGFs can strongly influence the corner frequency estimates for the target events, we focus here on optimal selection of EGFs and recording stations for a select number of moderate magnitude (Mw 4-5) earthquakes in the Ridgecrest sequence. We explore the EGF parameters which most influence the corner frequency estimates of these target events. By carefully analyzing the relationship between estimated corner frequency and EGF-specific parameters such as depth, time window length, cross correlation, high and low frequency bandlimits, and magnitude difference, we plan to increase the stability of and confidence in the corner frequency measurements. Finally, we will also compare target events with similar magnitudes but significantly different corner frequencies (and therefore very different stress drops) within the context of their EGFs to understand if the differences are real or artifacts of the EGF method.