The long-term seismic risk potential of active faults is generally estimated based on the characteristic earthquake model, which involves a fault being repeatedly ruptured with the same recurrence interval and the same displacement. However, in the last two decades, smaller amounts of displacement than the estimated slip based on the characteristic model have been observed along limited parts of active faults. Consequently, we must continue to investigate the magnitude and timing of earthquakes caused by faults in recent geologic time.

The Itoigawa–Shizuoka Tectonic Line active fault system (ISTLAFS) is considered one of Japan’s fault zones with the highest seismic risk. The timing of surface faulting events during the middle to late Holocene has been revealed for several sites along the northern and the north–central sections of the ISTLAFS. In contrast, the number of paleoseismic surveys conducted in the southern and the south–central sections of the ISTLAFS is limited.

In this study, we observed the sediments forming the fluvial terrace and conducted a trench excavation survey on the right bank of the Ojiro River on the Hakushu fault in the southern part of the ISTLAFS. The interpretation of the stereo images generated using LiDAR data allowed us to identify fault scarps beneath the forest and select the trench sites in which fine sediments were continuously deposited.

As a result, the vertical slip rate during the last ca. 9,000 years is estimated to be 0.56 mm/yr based on the formation age and vertical displacement of the fluvial terrace. The timing and amount of the coseismic slip of the paleoseismic events were estimated based on the displacement and deformation in the sedimentary units on the trench walls. The most recent and penultimate events are estimated to have occurred at 2,110–1,640 cal BP with a 1-m vertical slip and at 4,280–2,990 cal BP with a 0.5-m vertical slip, respectively. The interval between the two most recent activities is 880–2,640 years, which is approximately half of the recurrence interval mentioned in previous studies and almost the same as the elapsed time since the most recent faulting.

The earthquake recurrence models were verified using the two recent events and the slip rate on the Hakushu fault, as revealed in this study. This study’s time-slip diagram employs both the time-predictable and the slip-predictable models. According to the time-predictable model, the next faulting along the Hakushu fault will occur soon. Alternatively, based on the slip-predictable model, the vertical surface displacement of the next faulting is large (1 m), implying that the surface faulting rupture will occur along the southern section of the ISTLAFS.