Transient poroelastic stress coupling between the 2015 M7.8 Gorkha, Nepal earthquake and its M7.3 aftershock

The large M7.3 aftershock occurred 17 days after the 2015 M7.8 Gorkha earthquake. We Investigate if this sequence is mechanically favored by the mainshock via time-dependent fluid migration and pore pressure recovery. This study uses finite element models of fully-coupled poroelastic coseismic and postseismic behavior to simulate the evolving stress and pore-pressure fields. Using simulations of a reasonable permeability, the hypocenter was destabilized by an additional 0.15 MPa of Coulomb failure stress change (Delta CFS) and 0.17 MPa of pore pressure (Delta p), the latter of which induced lateral and upward diffusive fluid flow (up to 2.76 mm/day) in the aftershock region. The M7.3 location is predicted next to a local maximum of Delta p and a zone of positive Delta CFS northeast of Kathmandu. About 60% of the aftershocks occurred within zones having either Delta p > 0 or Delta CFS > 0. Particularly in the eastern flank of the epicentral area, similar to 83% of the aftershocks experienced postseismic fluid pressurization and similar to 88% of them broke out with positive pore pressure, which are discernibly more than those with positive Delta CFS (71%). The transient scalar field of fluid pressurization provides a good proxy to predict aftershock-prone areas in space and time, because it does not require extraction of an assumed vector field from transient stress tensor fields as is the case for Delta CFS calculations. A bulk permeability of 8.32 x 10(-18) m(2) is resolved to match the transient response and the timing of the M7.3 rupture which occurred at the peak of the Delta CFS time-series. This estimate is consistent with the existing power-law permeability-versusdepth models, suggesting an intermediately-fractured upper crust coherent with the local geology of the central Himalayas. The contribution of poroelastic triggering is verified against different poroelastic moduli and surface flow-pressure boundaries, suggesting that a poroelastic component is essential to account for the time interval separating the mainshock and the M7.3 aftershock.