Seismic velocity changes, strain rate and non-volcanic tremors during the 2009–2010 slow slip event in Guerrero, Mexico

Abstract : Seismic velocity changes, strain rate and non-volcanic tremors during the S U M M A R Y We use ambient noise cross-correlations to monitor small but reliable changes in seismic velocities and to analyse non-volcanic tremor (NVT) intensities during the slow slip event (SSE) that occurred in 2009 and 2010 in Guerrero. We test the sensitivity of the seismic velocity to strain variations in absence of strong motions. The 2009-2010 SSE presents a complex slip sequence with two subevents occurring in two different portions of the fault. From a seismic array of 59 seismometers, installed in small antennas, we detect a velocity drop with maximum amplitude at the time of the first subevent. We analyse the velocity change at different period bands and observe that the velocity perturbation associated with the SSE maximizes for periods longer than 12 s. Then a linearized inversion of the velocity change measured at different period bands is applied in order to determine the depth of the portion of the crust affected by this perturbation. No velocity change in the first 10 km is detected. Below, the velocity perturbation increases with depth, affecting the middle and lower crust. Finally, we compute the transient deformation produced by the SSE in an elastic model using the slip evolution recovered from the inversion of continuous GPS. The comparison between the velocity changes and the deformation suggests that the velocity change is correlated with the strain rate rather than with the strain. This result is similar to what was observed during the 2006 SSE in the same region and suggests a non-linear behaviour of the crust. The velocity changes can be interpreted together with other observables such as NVTs. During the 2009-2010 SSE we measure NVT activity using continuous seismic records filtered between 2 and 8 Hz. We observe a correlation between velocity changes (for period band greater than 14 s) and tremor activity whereas no correlation exists between velocity changes and seismic noise energy measured at long periods. These observations suggest that both seismic velocity change and NVT can be used as indication of transient deformation at depth.