Imaging afterslip along the Sumatran and Ecuadorian megathrusts over different timescales

Afterslip following large megathrust earthquakes represents an important postseismic process through which earthquake-induced stresses are released and redistributed. Mapping out the spatio-temporal distribution of afterslip can help us characterize seismic hazard along neighbouring portions of the fault, and provide insights into the mechanical properties of the megathrust. I will present case studies that depict the spatio-temporal distribution of afterslip following large earthquakes on the Sumatran and Ecuadorian megathrusts, over timescales that range from several minutes to a few years after the mainshock. In Sumatra, inversion of ~6.5-year-long GPS time series following the 2007 Mw 8.4 Bengkulu earthquake reveals that peak coseismic slip and afterslip areas are overall complementary. Our inversion technique is able to account for viscoelastic deformation following the earthquake. A key result from our study is that localized increases in stresses and strain rates caused by the coseismic slip and afterslip may have advanced rupture of a patch located updip of the 2007 rupture area, producing a Mw 7.8 tsunami earthquake in 2010. In Ecuador, inversion of high-rate GPS time series following the 2016 Mw 7.8 Pedernales earthquake reveals the signature of early afterslip in the initial minutes to hours after the mainshock, bringing us closer to the transition between the coseismic and postseismic phases. Our results suggest that early afterslip contributes substantially, ~ 60 %, to the postseismic geodetic moment over the first 3 days. The spatial signature of early afterslip in the shallow portion of the megathrust is consistent with longer-term afterslip over the first month, and spatially overlaps areas of reported and inferred slow slip events. These inversion results are guiding our exploration of frictional properties on the megathrust.