This paper examines the accuracy of estimating the distribution of damage to wooden houses and the accuracy of estimating road blockages caused by building collapse using incremental dynamic analysis (IDA) for Mashiki town, which was severely damaged by the 2016 Kumamoto earthquake aiming to accurate real-time earthquake damage estimation using basic survey of city planning. First, this paper estimates the distribution of the peak ground velocity by spatial interpolation method, and discussed the necessity of high-density seismic observation in areas where the peak velocity differs greatly over a short distance. Next, we focus on the difference between the records of observation at the ground surface and the records of observation on the first floor of the Mashiki Town Hall this paper estimates the damage to wooden houses in the area of Mashiki town against the mainshock by three types of curves: IDA curves statistically processed from multiple seismic waves observed in the past, IDA curves obtained from actual observation records at the ground surface, and IDA curves obtained from observation records on the first floor of the Mashiki Town Hall, respectively. As a result, IDA curves obtained from observation records on the first floor of the Mashiki Town Hall were the closest to the actual damage, suggesting the possibility of highly accurate real-time building damage estimation using the observed waves on the bottom floor of the building. Finally, the accuracy of the road closure method based on the estimated distribution of building damage was verified by comparing it with aerial photographs taken after the mainshock. Using this method, the actual damage result and the estimated results match 78.8% of the total length of road segments.

We conducted NIOM analysis for the earthquake records at the KiK-net Togi (ISKH04) site and examined nonlinear behavior of the ground at the site during the 2024 Noto Hanto earthquake. The results revealed that (1) the propagation time increased to 0.3136 s (NS) and 0.3245 s (EW) from the values before the mainshock of 0.2365 s and 0.2333 s, respectively; (2) propagation times at the last part of the mainshock records were still larger than those before the mainshock by 6 to 6.5 %; and (3) the SHAKE91 analysis indicated the significant nonlinear behavior in the soil layer from GL-28 to -50 m.