Current Research in EEPS Seminar: Allan Rubin – Princeton University
Abstract Title: “What is this thing called tremors?“
Nearly two decades after the discovery of tectonic tremor during slow slip in subduction zones, we still don’t have a good physical picture of what the tremor source looks like. Tremor is interpreted as consisting wholly or in large part of Low Frequency Earthquakes (LFEs), events with typical magnitudes of 1 to 2 but durations roughly 10 times longer than regular M1-2 earthquakes. Two end-member views are that (1) LFEs are fundamentally familiar, and represent slip on brittle patches in an otherwise creeping fault, similar to repeating earthquakes except that they take longer for a given magnitude, or (2) LFEs are an unfamiliar beast, and in fact are not really a “thing” at all; they just represent the stochastic accelerations/decelerations of a sliding fault that we see when focusing on the high-frequency limit of a continuous process (the “slow earthquake trend”) that extends down in frequency to month-long slow slip events.
We find that tremor seismograms of all amplitudes appear “saturated”, in the sense that they consist of one or more LFE arrivals per typical LFE duration (0.3 to 0.5 s), from the quietist tremor we see (comparable in amplitude to the noise on a tremor-free day) to the loudest (about 10 times larger). For a time-invariant magnitude-frequency distribution this would imply that loud tremor contains at least 100 arrivals per characteristic LFE duration, but some attributes of tremor are more easily explained if 10-times-louder tremor is made up of 10-times-larger LFEs rather than 100-times-more-frequent LFEs. I discuss what these observations and our recent tremor locations imply about the nature of the tremor source. Don’t expect a definitive answer.