| Summary: | During the last two decades, space geodesy allowed mapping accurately rupture areas, slip distribution, and seismic coupling by obtaining refined inversion models and greatly improving the study of great megathrust earthquakes. A better understanding of these phenomena involving large areas of hundreds of square kilometers came from the last gravity satellite mission that allowed detecting mass transfer through the Earth interior. In this work, we performed direct modeling of satellite GOCE (Gravity Field and Steady-State Ocean Circulation Explorer) derived gravity gradients up to degree/order N = 200 of the harmonic expansion and then corrected this by the effect of topography. Cutting off the model up to this degree/order allows inferring mass heterogeneities located at an approximate depth of 31 km, just along the plate interface where most (but not all) significant slip occurs. Then, we compared the vertical gravity gradient to well-constrained coseismic slip models for three of the last major earthquakes along the Sunda interface. We analyzed seismic rupture behavior for recent and for historical earthquakes along this subduction margin and the relationship of the degree of interseismic coupling using the gravity signal. From this, we found that strong slip patches occurred along minima gravity gradient lobes and that the maximum vertical displacements were related quantitatively to the gravity-derived signal. The degree of interseismic coupling also presents a good correspondence to the vertical gravity gradient, showing an inverse relationship, with low degrees of coupling over regions of relatively higher density. This along-strike segmentation of the gravity signal agrees with the along-strike seismic segmentation observed from recent and historical earthquakes. The thermally controlled down-dip ending of the locked fault zone along central Sumatra also presented an inverse relationship with the density structure along the forearc inferred using our modeling. From this work, we inferred different mass heterogeneities related to persistent tectonic features along the megathrust and along the marine forearc, which may control strain accumulation and release along the megathrust. Combining these data with geodetical and seismological data could possibly delimit and monitor areas with a higher potential seismic hazard around the world.
|
|---|