Study on Cumulative Activities of Passively Ruptured Faults through a Trenching Survey at the Matoishi Bokujo I Fault, Northwest Side of the Aso Caldera, Southwestern Japan

Abstract

Paleoseismic trenching surveys have helped estimate earthquake hazards based on the characteristic earthquake model, which does not consider passive rupturing. In this study, a trenching survey was carried out at the foot of the Matoishi Bokujo I fault scarp. The survey revealed cumulative vertical displacements of a passively ruptured fault reactivated by the 2016 Kumamoto earthquake (the 2016 earthquake; Mj 7.3). The Matoishi Bokujo I fault is located on the northwest outer rim area of the Aso caldera in Kumamoto prefecture, southwestern Japan, approximately 7 km from the Futagawa fault, which is the seismogenic fault of the 2016 earthquake. On the Matoishi Bokujo I fault where a fault scarp is clearly identified, a phase discontinuity caused by the 2016 earthquake was detected via displacement analysis using synthetic aperture radar interferometry of the Advanced Land Observing Satellite (ALOS)-2 data; the displacement measured 10-15 cm along the phase discontinuities. As aftershocks were not observed along the Matoishi Bokujo I fault, the phase discontinuities suggest that passive rupturing was induced by stress changes and/or strong shaking along the main Futagawa fault. On the trench wall of the uplift side of the fault scarp, an Aso-4 pyroclastic flow deposit ca. 1 m below the ground surface was covered with black soil; these deposits were deformed by normal faults. A clear displacement caused by the 2016 earthquake was not identified in the trench because it is assumed that the 10-15-cm displacement was separated into small displacements along the entire fault scarp. However, in the trench, two preceding faulting events at 3,430-2,890 cal BP and after 2,810 cal BP were identified with a displacement of up to 50 cm. This indicated the possibility that after 2,810 cal BP, the Matoishi Bokujo I fault could have passively ruptured simultaneously with the preceding earthquake on the Futagawa fault. In contrast, another earthquake triggered a passive or independent rupturing at 3,430-2,890 cal BP on the Matoishi Bokujo I fault. These results show that a fault scarp can be formed through different types of rupturing events. Although earthquake hazards related to active faults are typically evaluated based on the long-term slip rate estimated using tectonic landforms and recurrence intervals chronicled by trenching surveys, passively ruptured faults such as the Matoishi Bokujo I fault demonstrate the necessity of restructuring this evaluation method. To improve the method, distributions and characteristics of passively ruptured faults should be investigated in detail.