Detecting seismic signals and sources with networks of single stations and arrays

Abstract:

New, massive, datasets are being used to detect seismicity at lower magnitudes but require analytical methods that are both efficient and capable of extracting useful information from faint signals immersed in noise. We have developed the AELUMA (Automated Event Location Using a Mesh of Arrays) method that recasts any network of single sensors as a distributed mesh of triangular arrays (triads). Each triad provides a local estimate of signal properties. In the latest version of the method this information from triads across the network is combined with results from any arrays within the network footprint to estimate the source origin time and location. The process is repeated without oversight to catalog events.

A key challenge in attributing signals to their source occurs when a large number of signals are detected nearly concurrently from different sources. We apply a cluster (decision tree) analysis that takes the results of array processing at all triads and arrays to iteratively parse out subsets of detections from distinct sources.

We used recordings of seismic signals made at an extensive network of sensors to develop the hybrid method and build a catalog of seismic activity in the western United States. The accuracy of AELUMA is assessed using events for which the origin time and location are well known. The method has been further adapted for application to the relatively sparse IMS.