Imaging of seismic sources by surface-wave time-reversal: long-period earthquakes

The study of long-period events in a volcanic setting is of fundamental importance to better understand the physics of volcanic plumbing systems. We locate such events using a source-imaging method developed by our team, and successfully applied, e.g., to the great Sumatra earthquake (Dhakal et al. 2022). Our approach combines seismic time reversal with a surface-wave ray tracing algorithm based on generalized spherical-harmonic parameterization of surface-wave phase velocity, and accounting for azimuthal anisotropy. We present a new application, to recordings of a suite of Mayotte events that Cesca et al. (2020) have already studied and interpreted in terms of the drainage of a magma reservoir.

We first conduct synthetic tests to quantify the resolving power of our method, given the available data coverage for the events of interest. We then use low-frequency Rayleigh wave signals recorded by different stations, reverse them in time and back propagate them through a surface-wave phase-velocity model. The time-reversed wave field has a prominent maximum at the spatial location(s) and time(s) where and when the recorded signal had been generated. From the time- and space-distribution of such maximum, we can make inferences on the nature of the source. Results so obtained are compared with those determined by Cesca et al. (2020) via moment tensor inversion and found to be in good agreement. We infer that our methodology is applicable to volcanic settings, possibly providing new insights into the nature of long-period seismic sources related to volcanic activity. The precise location of such events can provide better constraints on the depth interpretations and the extent of the seismic source.