An anomalous swarm of low-frequency events at Mount Baker, Washington, June-August 2009

Between June 16 and August 22, 2009, a swarm of least 39 low-frequency (LF) seismic events occurred at shallow depths beneath Mount Baker, a Cascade Range composite stratovolcano located in northern Washington. The LF events had several characteristics similar to classic volcanic low-frequency events, including narrowly peaked spectra (~3 Hz) and somewhat extended codas. They occurred at relatively regular time intervals, at first averaging one every 3-4 days and then accelerating in mid-July to one per day before tailing off in late August. Many were recorded at stations as distant as 220 km, and those that were located by the Pacific Northwest Seismic Network (PNSN) had coda-duration magnitudes of Md 1.5-2.2. Many events had similar (cross-correlation coefficient = 0.7) waveforms, indicating that the source type and location were similar for most events. A search through the PNSN catalog from 2000-2009 found 8 LF events with similar characteristics to the 2009 events; thus shallow LF events are not unprecedented at Mount Baker, but also clearly have never been detected at rates as high as the 2009 swarm. Mount Baker has been seismically monitored since 1972, although the network has consisted of only a single short-period seismometer ~6 km west of the summit and a total of 6 weak-motion seismometers within 50 km. The location threshold at Mount Baker is thus quite poor (estimated to be Md 1.6), and it is likely that some prior shallow LF events have gone undetected. Detection of the 2009 LF events was enabled in part by addition of 2 new PNSN stations at 30 and 38 km from Mount Baker (although these stations only improved the estimated location threshold from Md 1.7 to Md 1.6) and also through automatically produced, multi-station, 10-minute spectrograms that have made visual detection of such events much easier. Thus it is difficult to know whether the 2009 swarm is unprecedented at Mount Baker, although we can be reasonably sure that no similar swarms have occurred since installation of the new PNSN stations in 2006-2007. LF events with similar characteristics to the Baker events are commonly seen at restless volcanoes around the world and are usually attributed to magmatic or hydrothermal processes, such as pressurized fluids or gas moving through cracks in a conduit system. Mount Baker has an active hydrothermal system with a persistent magmatic-gas-rich plume emanating from a near-summit geothermal area, so such a source mechanism is not unreasonable. However, Mount Baker is a glacier-clad volcano, initial locations of the LF events by the PNSN place many near or beneath glaciers, and the timing of the swarm coincides with the maximum melt season. In addition, similar events have been attributed to glacier motion at other volcanoes. Thus a glacier-related source is also a viable hypothesis. In this paper we will discuss different arguments for various source models for the 2009 Mount Baker LF event swarm.