Using waveform correlation and template event metadata to reduce analyst workload.

Abstract

DISCUSSION : Using waveform correlation only as a signal detector does not guarantee a reduction in analyst workload if marginal events that do not meet bulletin-inclusion criteria are detected. Our research targets the selection of waveform correlation detections that will lead to bulletin-worthy events. The chosen example shows that a credible repeating event can be detected by dynamically creating additional templates from arrivals that were associated to the template event. The individual waveform correlation detections may not be convincing (for example, view the detections for ELK Pn and NVAR Pn in Step 5), yet the combined group of corroborating arrivals are consistent in time with a repeating event. The table above shows that a simple calculation of time deltas can establish the credibility of the repeating event and can establish a relative location to the template event. This research is in early stages, and the method will be applied to additional areas of high event rates of repeating events such as mining blast regions in Wyoming and Scandinavia and two aftershock sequences in the Middle East. At this time, only high time-bandwidth phases are used as templates (i.e., excludes teleseismic P templates). The research seeks to quantify parameters from the initial waveform correlation detection that favor the processing of templates for corroborating arrivals.