Microseismicity detection across the Antarctic continent

Though the Antarctic continent has been vastly unexplored due to obvious constraints, the last decade of scientific investment has yielded numerous regional- and continent-scale seismological experiments carried out by researchers. The public archiving of those data at the IRIS Data Management Center makes unintended discoveries possible. As an example, we utilize broadband seismic data from a recent experiment (TAMNNET), which was originally proposed as a structural seismology experiment across the Transantarctic Mountain area of Victoria and Oates Land, for detecting local seismic events. Most of the seismicity emanates from David Glacier, upstream of the Drygalski Ice Tongue, which has been documented by several other studies; however, additional areas include clusters of activity. In order to improve the catalog completeness for the David Glacier area, we utilize a matched-filter technique (also known as template matching) to detect potential missing earthquakes that may not have been originally detected. We expand this analysis and utilize existing phase picks for machine learning applications and compare the relative effectiveness between template matching and machine learning approaches. We have also downloaded other regional- and continental-scale seismological datasets from Antarctica and preliminary results suggest abundant small-magnitude seismicity that may illuminate interesting ice-dynamic, tectonic, or a combination of both processes. We plan to use machine-learning techniques to automatically classify different types of seismic events recorded around the Antarctic continent. Updated results will be presented at the meeting.