Revisiting the Source Rupture Process of the Mainshock of the 2016 Kumamoto Earthquake and Implications for the Generation of Near-Fault Ground Motions and Forward-Directivity Pulse

Abstract

Strong near‐fault ground motions associated with the M[JMA] 7.3 mainshock of the 2016 Kumamoto, central Kyushu, Japan, earthquake sequence have received attention by seismological and engineering communities. In this study, the kinematic source rupture process was reanalyzed based on an improved approach for the representation of source faults. The slips at densely distributed point sources were defined via the bilinear interpolation of those at surrounding control points. The result shows that the rupture started on the Hinagu fault with a small initial rupture and propagated beyond the junction to the Futagawa fault. The rupture on the Futagawa fault mainly propagated up and northeastward. A large slip area with a peak slip of 4.9 m and peak slip velocity of 3.1 m/s was detected at depths ranging from 3 to 15 km in the central part of the Futagawa fault. This asperity spatially coincides with a body with moderate‐seismic velocity (⁠VP∼6 km/s⁠) and low‐seismic attenuation. The slips on the Futagawa fault have significant normal‐slip components, whereas the slip vectors of the Hinagu fault represent almost pure right‐lateral strike slip. The shallower part of the fault segments in the western Aso caldera is characterized by relatively large normal slips. The estimated slip‐velocity functions at shallower depths (⁠<3 km⁠) are almost temporally symmetric and relatively long. The shallower portion of the source fault significantly contributes to the velocity and displacement waveforms at near‐fault sites. On the contrary, the slip‐velocity functions at deeper depths (⁠>3 km⁠) are temporally asymmetric and have a sharp peak. The simulation of the ground‐motion evolution suggests that the lateral flow in the Aso Valley was primarily triggered by the strong forward‐directivity pulse generated from the asperity on the Futagawa fault.