Isabelle Panet, Clément Narteau, Jean-Michel Lemoine, Marie Bouih, Sylvain Bonvalot, and Dominique Remy
Retrieving short-term preseismic signals before the occurrence of great subduction earthquakes is a major goal for seismic hazard mitigation. It requires a continuous monitoring of seismic and aseimic deformations within the entire subduction system, which cannot be obtained from space geodetic measurements of ground deformations and seismology only. With its unique sensitivity to mass redistributions at all depths along and around the descending slab, satellite gravity can bring key information to overcome this observational gap.
Taking the example of the 2011 Tohoku earthquake, we develop here a global and systematic analysis of time series of GRACE-reconstructed gravity gradients, to test whether the deep gravity signals preceeding the earthquake (Panet et al., 2018 ; Bouih et al. 2022) can be detected before the occurrence of the event, as a specific feature originating from solid Earth. Our approach is based on a refined detection of short-term gravity variations aligned with the subduction direction, and consistency tests between signals shared by different GRACE geoid models, allowing us to evaluate the signals robustness with respect to errors and oceanic corrections in the GRACE gravity solutions. Considering the GRGS and the CSR solutions, the results of these tests underline the unique character of the deep pre-Tohoku gravity variations compared to the usual variability of gravity, confirming its interpretation as pre-seismic signal.
We compare the preseismic gravity signals of the 2011 Tohoku and 2010 Maule earthquakes, and conclude that our approach can be applied to the systematic monitoring of the Pacific subduction belt.