Does the 2007 Noto Hanto earthquake reveal a weakness in the Japanese national seismic hazard map that could be remedied with geological data

The Noto Hanto earthquake struck one of the lowest earthquake probability regions on the national seismic hazard map of Japan. To contribute to future updates of the hazard map, we examined the predictability of the 2007 earthquake on the basis of geological data that were available before it occurred. Sonic prospecting profiles of active faulting and the absence of an onshore fault could have limited the potential rupture length to 12–15 km, similar to the 2007 source. Empirical relationships between magnitude and fault length would have given us Mj = 6.6–6.8 and Mw = 6.3–6.4. The emergence of one marine terrace, which inclines to the south and reaches an altitude of approximately 50 m, can be dated to 120–130 ka and yields an uplift rate of approximately 0.4 mm/year. Mw-displacement empirical relationships and examples of recent blind fault events that have occurred at various locations around the world suggest that the conceivable maximum coseismic uplift of such shocks is 40–70 cm. Together with the uplift rate, we would have obtained an average recurrence interval of 1000–2000 years and, consequently, a 1.5–3.0% time-independent (Poisson) probability for 30 years. In addition, the significant inclination of the marine terraces—3.2 per mille (0.32%)—is better explained by the accumulation of frequent southward tilting as large as that of the 2007 type event with approximately 1600-year intervals, without any significant contributions from other seismic sources. We therefore conclude that the Noto Hanto earthquake source would have been better evaluated and identified if we had taken into account not only major active faults but also the active tectonics of moderate-size faults and their associated scale and rate.