Spatio-temporal variations of afterslip and viscoelastic relaxation following the Mw 7.8 Gorkha (Nepal) earthquake

Abstract

We use 4 years of 3D GPS data from Nepal and southern Tibet to investigate the postseismic deformation caused by the 2015 Mw 7.8 Gorkha (Nepal) earthquake. We first model afterslip and viscoelastic relaxation separately, but find that this approach results in an overestimate of the total postseismic deformation. We then use an integrated model to simultaneously extract the contributions from afterslip and viscoelastic relaxation by assessing the misfit between observed and simulated displacements during different periods. The results show that the near-field postseismic displacements are dominated by downdip afterslip during the first 2 years, and then viscoelastic relaxation plays the leading role in the following years. In the far-field, however, the observed deformation is mainly controlled by viscoelastic mechanism throughout the postseismic period. The best model supports a laterally heterogeneous rheology: the Tibetan lower crust is viscoelastic with a transient viscosity 5×1017 Pas, steady-state viscosity 5×1018 Pas; we assume in our model that India has an elastic lithosphere 50 km thick with a high viscosity upper mantle (1020 Pa s), and that the same mantle viscosity applies beneath Tibet. We further predict that viscoelastic deformation will be observed in the near-field for ∼12 years and the displacements caused by afterslip for ∼6 years. Given the afterslip following the earthquake and the possible untapped strain inherited from historical events, the regions to the west and south of Kathmandu have the potential to suffer a large earthquake rupture in the future.