Inversion of the spatiotemporal distribution of early postseismic slip within the rate-and-state framework using high-rate GNSS position time series
Résumé
Whereas the spatial and temporal evolution of early postseismic slip (i.e., the first few hours) has been shown to be complex, we do not know well the mechanisms that control its behaviour. One pending question is to know whether or not the rate-and-state friction law is required to explain it, or if the rate-dependent friction law, traditionally used to explain afterslip observations on a day-to-month time scale, is sufficient. Based on 30-sec high-rate position time series that show the early postseismic surface displacements at 13 stations from 2.5 minutes to 3 days after the 2015 M8.3 Illapel, Chile, earthquake, we attempt to infer the frictional properties that control early postseismic slip on the fault. We invert simultaneously for the location and the amplitude of the slip patches, and for the frictional parameters that control their temporal behaviour. We use the elliptic slip patch inversion method (Vallée et Bouchon, 2004; Twardzik et al., 2012) that allows to reduce the number of unknown parameters and we solve, in particular, for the rate-and-state friction law parameters. The obtained frictional parameters are consistent with a velocity-strengthening behaviour. Our first results indicate that 1) the framework can explain the full time series, 2) the postseismic slip patches are located on the edge of the coseismic slip distribution, 3) the obtained stress drop parameter is consistent with that found for the coseismic rupture, and 4) the stiffness of the rate-and-state friction law is consistent with the size of the slip patches. We will also discuss the ability of compatible models to apply at longer postseismic timescales which would suggest a unique underlying physical process.