Evaluating event detection success in the Alaskan Arctic using hydrophone data

Ocean-based monitoring complements land-based seismoacoustic networks by adding monitoring capacity to the 70% of earth’s surface covered by water. Ocean-based stations can reduce detection time and increase azimuthal coverage for coastal and offshore events and provide records of T-phases not always recorded at land-based stations. Events recorded at ocean-based instruments must compete with the ocean’s ambient soundscape, which, at frequencies used for earthquake detection (1–100 Hz), includes ocean waves, sea ice activity, and anthropogenic sounds like seismic airgun surveys. Here we evaluate detection success for known submarine earthquakes at a NOAA Ocean Noise Reference Station (NRS) hydrophone located north of Point Barrow, Alaska. We determine a baseline detection threshold for these events with noise levels typically recorded at the hydrophone. Additionally, we use a short-time-average over long-time-average (STA/LTA) detector to identify additional events not present in earthquake catalogs and demonstrate that hydroacoustic records can improve the catalog for local-to-regional submarine events in the sparsely monitored Arctic. Knowing that the soundscape is actively changing in the Arctic due to decreasing sea ice and increasing human activity, among other shifts, we demonstrate a noise injection method to investigate the impacts of expected future sound levels on event detection thresholds.