Coseismic changes of gravitational potential energy induced by global earthquakes based on spherical‐Earth elastic dislocation theory

We compute the coseismic gravitational potential energy E g change using the spherical‐Earth elastic dislocation theory and either the fault model treated as a point source or the finite fault model. The rate of the accumulative E g loss produced by historical earthquakes from 1976 to 2016 (about 42,000 events) using the Global Centroid Moment Tensor Solution catalogue is estimated to be on the order of −2.1 × 10 20  J/a, or −6.7 TW (1 TW = 10 12  W), amounting to 15% in the total terrestrial heat flow. The energy loss is dominated by the thrust faulting, especially the megathrust earthquakes such as the 2004 Sumatra earthquake ( M w  9.0) and the 2011 Tohoku‐Oki earthquake ( M w  9.1). It is notable that the very deep focus events, the 1994 Bolivia earthquake ( M w  8.2) and the 2013 Okhotsk earthquake ( M w  8.3), produced significant overall coseismic E g gain according to our calculation. The accumulative coseismic E g is mainly lost in the mantle of the Earth and also lost in the core of the Earth but with a relatively smaller magnitude. By contrast, the crust of the Earth gains gravitational potential energy cumulatively because of the coseismic deformations. We further investigate the tectonic signature in the coseismic crustal E g changes in some complex tectonic zone, such as Taiwan region and the northeastern margin of the Tibetan Plateau. We found that the coseismic E g change is consistent with the regional tectonic character.