A new automated approach to detecting and locating seismic events using data from a large network

TitleA new automated approach to detecting and locating seismic events using data from a large network
Publication TypeJournal Article
Year of Publication2018
Authorsde Groot-Hedlin C.D, Hedlin MAH
JournalBulletin of the Seismological Society of America
Volume108
Pagination2032-2045
Date Published2018/08
Type of ArticleArticle
ISBN Number0037-1106
Accession NumberWOS:000440400000016
Keywordsarray; central oklahoma; earthquakes; Geochemistry & Geophysics; injection; phase pickers; united-states
Abstract

The Automated Event Location Using a Mesh of Arrays (AELUMA) method, originally developed for detection of atmospheric sources using infrasonic data, is modified here to detect and locate seismic events. A key feature of AELUMA is that it does not require a detailed velocity model to locate events. The new method was applied to vertical-component seismic data recorded by the USArray Transportable Array to (1) test its efficacy when applied to a very large dataset, (2) test its ability to detect and accurately locate distinct event types across a geologically diverse region without analyst oversight, and (3) assess the sensitivity and accuracy of the method. Using data filtered from 1 to 8 Hz, 9996 events were detected in clusters within the central United States-with most events located near areas known for anthropogenic activity. The method was compared with three catalogs in Oklahoma-a region known for small anthropogenic events. In comparison with accurate locations from a template study, AELUMA detected all events from M-L >= 1.9 but none below M-L 1.3. The median absolute origin time and location offset were 9.5 s and 6.0 km, respectively. Comparisons of AELUMA's catalog in Oklahoma with two others (produced by Oklahoma Geological Survey [OGS] and the Array Network Facility) showed that AELUMA found more events than either catalog, including clusters of emergent events that were largely missed by the other methods. However, most of the smaller magnitude events detected by OGS were missed by AELUMA, mainly due to the sparser network used by AELUMA.

DOI10.1785/0120180072
Student Publication: 
No
sharknado