Gradients from GOCE reveal gravity changes before Pisagua Mw = 8.2 and Iquique Mw = 7.7 large megathrust earthquakes

Considerable improvements in the measurement of the Earth gravity field from GOCE satellite mission have provided global gravity field models with homogeneous coverage, high precision and good spatial resolution. In particular, the vertical gravity gradient (Tzz), in comparison to the classic Bouguer anomaly, defines more accurately superficial mass heterogeneities. Moreover, the correction of these satellitederived data from the effect of Earth topographic masses by means of new techniques taking into account the Earth curvature, improves results in regional analyses. In a recent work we found a correlation between Tzz and slip distribution for the 2010 Maule Mw= 8.8 earthquake. In the present work, we derive the vertical gravity gradient from the last GOCE only model, corrected by the topographic effect and also by the sediments on depocenters of the offshore region at the PerueChile margin, in order to study a spatial relationship between different lobes of the gravity derived signal and the seismic sources of large megathrust earthquakes. In particular, we analyze this relation for the slip models of the 1996 Mw = 7.7 Nazca, 2001 Mw = 8.4 Arequipa, 2007 Mw = 8.0 Pisco events and for the slip models of the 2014 Mw = 8.2 Pisagua and Mw = 7.7 Iquique earthquakes from Schurr et al. (2014), including the previously analyzed 2010 Mw = 8.8 Maule event. Then we find a good correlation between vertical gravity gradients and main rupture zones, correlation that becomes even stronger as the event magnitude increases. Besides this, a gravity fall in the gravity gradient was noticed over the area of the main slip patches at least for the two years before 2014 Mw = 8.2 Pisagua and Mw = 7.7 Iquique earthquakes. Additionally, we found temporal variations of the gravity field after 2010 Mw = 8.8 Maule event, related to the main patches of the slip distribution, and coseismic deformation. Therefore, we analyzed vertical gravity gradient field variations as an indirect measure of the variable seismic coupling finding a potential relationship between Tzz and the seismic b-value. These relationships exemplify the strong potential of the satellite only derived models as a predictive tool to determine potential seismic energy released in a subduction segment, determining the potential size of a potential rupture zone, and in particular internal slip distribution that allows inferring coseismic displacement field at surface.