A Generalized Brune Model to Identify Non-Earthquakes and to Explore Earthquake Scaling

Abstract:

The Brune (1970) model of an idealized point-source earthquake S-wave spectrum has proven extremely useful for many seismological applications. Here we discuss an updated and more generalized version of this earthquake spectra model and its application to identifying non-earthquake events and studying earthquake scaling. The generalized model includes both P and S-waves and allows variable apparent stress to calculate predicted spectra for events over a broad region based only on their size and location (Walter and Taylor, 2001). This Magnitude and Distance Amplitude Correction (MDAC) can be used to normalize observed spectra and flag anamoulous events, helping identify explosions and cavity collapses. The MDAC earthquake spectra are based on the idealized Brune spectral shape and uses energy conservation to tie back to the source (e.g. Walter and Brune, 1993). By normalizing observed spectra in this way explosions, mined cavity collpases and other non-earthquakes events stand out when using ratios of P and S-wave amplitudes at various frequencies (e.g. Walter et al., 2018). Here we show how the updated Brune spectral model was developed for MDAC and apply it to example earthquakes, explosions and collapses in the western U.S. and Eastern Asia. We also show how in the process of using earthquakes in a region to calibrate the MDAC parameters, the best fits in many regions seem to show apparent stress is non-constant and appears to be larger for M>5 events than for smaller events (e.g. Walter et al., 2006). One challenge is getting an accurate azimuthally averaged estimate of the source spectra. We have been successfully using the scattered, late-arriving coda energy to do this via regional envelope fitting (e.g. Mayeda et al., 2003). We are making new open-source tools available to do this coda derived earthquake source spectral modeling and encourage interested users to try them (see: https://github.com/LLNL/coda-calibration-tool). We are working to expand functionality to include coda enevelope ratio analysis (e.g. Walter at al., 2016) to better reduce the often confounding effects of path and site in determination of source parameters such as corner frequency and apparent stress.