The inhomogeneity of the medium in the seismic propagation path is inferred to vary depending on regions and depth. To improve the accuracy of numerical simulations of earthquake motion in the random media adequate inhomogeneity parameters consistent with the region where the seismic waves propagate are needed. In this study inland earthquakes are focused on and strength(ε) and correlation length(a) of inhomogeneity distribution of seismic velocity in the propagation path are inferred. First ε²/a values of inhomogeneity of seismic velocity in the propagation path were inferred based on KiK-net observation records whose hypocentral distances are between 10km and 60km. In order to estimateε²/a least square regression analysis was performed in which fitting of the envelope time series of observed records and the calculated envelope based on scattering theories. It was revealed that the estimated values of ε²/a from the short-distance (about 20km) records were greater than values in the previous studies. Then numerical simulations were performed in order to study the influence of the inhomogeneity on seismic wave. Three different 3D random inhomogeneous models were set up and the finite difference method was applied to obtain velocity time histories in the random media. As a result dependency of ε²/a on the hypocentral distance similar to the observed records has been confirmed. In the analyzed distance range the simulation result based on the inhomogeneity parameters of ε=0.05 and a=4km has relatively agreed with the result from the observed records.

This paper proposes a methodology to apply Bayesian inference in post-earthquake assessment to estimate damaged houses distribution. It uses a damage inspection report on limited areas as ’observation’ in Bayesian protocol for updating the error in parameters of fragility functions. A numerical study with the 2011 off the Pacific coast of Tohoku Earthquake record is also presented so as to examine the characteristics of the presented method.

An activity time of DMAT for the life-saving treatment during the acute phase is less than 3 days from an earthquake occurrence. Therefore, DMAT have to arrive at disaster stricken area immediately but an air route and a land route as the means of transportation have a high possibility of simultaneous damage. There is also delay of gathering of the commander, medical appliances and disaster information. In this paper, the method of evaluating the efficiency of DMAT is proposed in consideration of the delay of gathering of the personnel required and the damage of the land route and air route facilities. We applied the proposed method and evaluated fatalities to the medical unit of a certain prefecture as an example. In addition, we evaluated the decrease of fatalities by earthquake-resistant measures.

The Tohoku Earthquake of March 11, 2011 was associated with widespread liquefaction-induced damage in the eastern part of Japan. Our continuing surveys nearly five years from soon after the earthquake revealed that the earthquake induced liquefaction at a significant number of locations in a 650-km long zone extending in the eastern part of Japan. This paper summarized the characteristics of liquefaction during the earthquake. Approximately 8600 grid cells identified to contain liquefied sites; the size of each grid cell was approximately 250 m × 250 m. The intensities of the ground motion of the liquefied grid cells were examined, and approximately 95 % of liquefied grid cells were found to be distributed in the areas where the I_JMA exceeded 5 upper during the main shock. Approximately 98% and 99% of liquefied grid cells were distributed in the areas where PGAs exceed 140cm/s² or where PGVs exceed 15cm/s, respectively. The liquefaction occurrence ratios (number of liquefaction grid cells divided by total number of grid cells) of different geomorphological unit were calculated where I_JMA exceed 5 upper. Those for artificial fills, sand dune, former river channel are large in this order. There was much difference between Tohoku and Kanto regions both in number of liquefaction and liquefaction occurrence ratios. To reveal the cause of the differences between two regions, the factor of safety (F_L) against liquefaction was evaluated at 231 sites where liquefaction occurred during the 2011 earthquake. Furthermore, likelihood of liquefaction-induced damage was evaluated based on the combination of liquefaction potential index P_L and thickness of surface un-liquefiable layer, H₁ overlying liquefiable layers, which were determined from the result of the F_L analysis. Comparing the result of these analyses of two regions, it was found that the ground of Kanto region is much more likely to liquefy than Tohoku region.

In this study, we constructed disaster risk communication method on disaster prevention activity, contributed to the improvement of local disaster prevention capability. The method takes into consideration three steps: estimation of damage, investigation of countermeasures, and activities and verification. The effect of the method was investigated through activities in an elementary school district in Tsukuba City. As a result of cooperation of various local communities and risk communication while evaluating the inherent local risks and investigating countermeasures, roles and stakes of various local bodies with respect to disaster risk and the necessity of a contact system were confirmed. Based on these findings, by reviewing the existing disaster prevention system it was possible to construct a staged dispersed evacuation system and cooperating networks that included sharing information and supplies within the school district, etc., as well as to devise disaster prevention measures using social resources.

A large response to ground motion was predicted in a seismically isolated building at a near-fault location. The response distribution of isolated buildings was estimated by utilizing the disclosed ground motion records from the 2016 Kumamoto earthquake, as observed by K-NET and KiK-net. The results showed the existence of a limited local area near the end of a fault rupture where severe displacements occurred in the seismic isolators. The area was near the end of a fault rupture. The observed displacement of the isolators of the actual buildings agreed with these estimated values.