PNNL

Abstract

Abstract: We have developed and tested a unifying algorithm for estimating the location of a seismic event. The estimation approach differs from established non-linear regression techniques by using Bayesian priors to incorporate ancillary physical basis constraints and knowledge about the event location. P-wave (primary) arrival times from seismic data waveforms are used in the development. Depth, a focus of this paper, may be modeled with a probability density function (potentially skewed) that captures physical basis bounds on depth from ancillary event characteristics. For instance, the surface wave Rg is present in a waveform only when an event is shallow. A high confidence Rg detection in one or more event waveforms can lead one to assume a shallow-skewed prior probability density function for the depth parameters. The proposed Bayesian algorithm is illustrated with a magnitude 5.6 earthquake in southwest Montana by comparing results with good and poor station configurations. A noninformative uniform prior is used in both cases to demonstrate a hypocenter estimate that is equivalent to the established non-linear regression approach.