In this paper, we estimated the influence of slip velocity function on the reproducibility of observation records by the theoretical method. First, we collected source fault models based on the waveform inversion of strong motion data (hereinafter called "Inversion Model"). In addition, we constructed characterized source models based on strong ground motion prediction method by the Headquarters for Earthquake Research Promotion (hereinafter called "Recipe Model") . Next, we simulated observation records using these source fault models by the theoretical method. It was found that the calculated waves using Recipe Model considering the region shallower than seismogenic layer overestimated observation records near source region. These results were caused by difference of the slip velocity functions in the region shallower than the seismogenic layer between Inversion Model and Recipe Model. Based on the above results, we recommended the Yoffe-type slip velocity functions modified by Tinti et al. (2005) for shallower region than the seismogenic layer from Inversion Model.

In the previous study, a 98-dimensional feature vector was proposed to represent temporal characteristics of a waveform of ground motion acceleration on the basis of its Husid plot. In this study, a method for mode decomposition and synthesis of the temporal characteristics was proposed using Karhunen-Loéve (KL) expansion. First, eigenvalues and eigenvectors are solved for variance-covariance matrix defined for the 98-dimensional feature vectors calculated using a number of accelerograms. Next, the feature vectors are transformed to principal components using the orthogonal matrix composed of the eigenvectors. Furthermore, inverse transformation is applied to several most principal components constituting the subspace in order to reduce the dimension of the original feature vectors with allowable loss of information. The kernel density estimation method is used to validate the envelope recomposed by mode synthesis. Illustrative examples for acceleration records observed in the 2011 Off the Pacific Coast of Tohoku Earthquake, Japan are shown. It is shown that three to six dimensional representation of the temporal characteristics can be generally achieved by using the orthonormal basis extracted by the KL expansion. On the other hand, principal components of higher order are required to describe the temporal characteristics of waveforms whose amplitude levels contain much fluctuations.

We investigated the distribution of the collapsed buildings due to the 2016 Kumamoto earthquakes in the area along the Akitsu river. A Mj6.5 foreshock and Mj7.3 mainshock occurred within 28 hours. The spatial pattern of collapsed buildings from the foreshock and mainshock were similar and the isolated areas resemble islands. Since the surface rupture occurred only during the mainshock, it is unlikely that the presence of the surface rupture generated the similar damage patterns for the foreshock and mainshock in Mashiki. The most severely damaged areas are located in the lowest river terrace. The isolated islands correspond to the locations of built areas constructed during the Meiji era. The cause of the damage islands is likely to be a combination of the subsurface soil structure and age of buildings.

This study tried to detect collapsed bridges due to tsunami by analyzing changes of backscattering intensity in Synthetic Aperture Radar (SAR) images. TerraSAR-X images covering the Sendai and Ishinomaki planes acquired before and after the 2011 Tohoku-Oki earthquake were employed together with the GIS data of bridges. The bridge region was created according the GIS data with a 10-m buffer. Then the difference and correlation coefficient of the sigma naught values within the bridge region were obtained. The proper threshold value of the correlation coefficient was investigated to extract collapsed bridges. In order to verify the effectiveness of the method, the extracted results were compared with the reference data from damage investigation reports and aerial photographs. The threshold vale of the correlation coefficient, 0.47, gave the highest value for the summation of the producer accuracies for collapsed and survived bridges, with the overall accuracy 0.776 and the kappa coefficient 0.430.

We developed a new method to calculate displacement of elastic half-space due to slip of the source fault using Boundary Element Method (BEM). The fault having irregular shape or irregular slip distribution can be modeled continuously to make use of triangular liner element which is usually applied by BEM. However, BEM has a disadvantage that the displacement is usually calculated unstable at an observation point that is near an element due to singularity of Green's function. On the other hand, the displacement is obtained exactly by the developed method to exclude the singularity using PART method by Hayami and Brebbia (1988). For the greatest earthquake that is expected in the future to break out in the Sagami Trough, we calculate the displacement of the seabed by the developed method and by Okada's solution (1985) at first. Next, we perform tsunami simulation using these displacements that are regarded as the initial water levels. Since the displacement calculated by the developed method does not contain the effect of modeling the fault discontinuously which appears in Okada's solution, the tsunami simulation can be executed stably.

In order to estimate shear-wave velocity (Vs) structure model at an earthquake motion observation site, a joint inversion technique using observed earthquake motion data and microtremor exploration data is developed. Site amplification factor, receiver function extracted from earthquake motion data and phase velocity of Rayleigh wave extracted from microtremor array observation in Hofu City, Yamaguchi Prefecture are analyzed in this study. Vs structure models up to the seismic bedrock are estimated in the inversion by minimizing the total misfit between the observed and the theoretical data. The Vs structure models that reproduce the observed characteristics are successfully obtained by the proposed joint inversion technique. In order to identify the two-dimensional subsurface structure in Hofu City, the inversion technique is applied to five observation sites by assuming the unified Vs for deep underground model calibrated from the preliminary inversion analysis. The result shows that thickness of the sediment in the central part of the basin is larger than those in other sites, indicating the significant irregularity in the basin.

Database of super high-rise residential buildings in Japan were compiled to grasp their spatial and temporal variation. In addition, extensive response analysis were performed for RC super high-rise residential buildings during huge earthquakes including the 2011 off the Pacific Coast of Tohoku earthquake and the 2016 Kumamoto earthquake. Spatial variation of peak inter-story drift angle (PIDA) of buildings and relationship with response spectra levels was discussed for rapid estimation of building damage. Finally, spatial distribution of PIDA was evaluated using input earthquake motions for the structural design of long-period buildings due to the hypothetical Nankai Trough Earthquake provided by MLIT, showing the validity of the proposed regression equation.

Spectral decay characteristics in high frequency range due to large and small inland crustal earthquakes that occurred in Japan are examined. The Butterworth type high-cut filter with cut-off frequency, f_max and its power coefficient of high-frequency decay is assumed in this study. Four parameters such as seismic moment M_o, corner frequency f_c, f_max, and its power coefficient s are estimated by comparing observed spectra at hard rock sites with theoretical spectra. The theoretical spectra are calculated, based on the omega squared source characteristics convolved with propagation-path effects and f_max filter shapes. In result, f_max's of six large earthquakes are estimated to be 7Hz to 10Hz and those of small earthquakes are estimated to be 9Hz to 20Hz. The power coefficients of the f_max filter, s, for earthquakes in Shizuoka and Kumamoto regions are larger than those for earthquakes in other regions. Therefore, a region dependency of spectral decay characteristic is suggested. Moreover, correction filter, P_C(f), a filter to correct the difference of spectral decay characteristics between large and small earthquakes, are suggested. Finally, effects of spectral decay characteristics from large earthquakes on strong ground motion prediction are examined. The results in this study contribute to strong ground motion prediction in high frequency range.

A large-scale landslide in proximity of Aso Bridge, which is located near the epicenter of the main shock of the 2016 Kumamoto Earthquake were occurred due to the strong motion. In this study, observations for microtremor and aftershock were conducted at the target site in order to evaluate the empirical site effects. Seismic waveforms at the landslide site during the 2016 main shock were then estimated based on the site effects and the characterized source model. The estimated ground motions will be useful in the detailed study of the landslide mechanism.

During the 2016 Kumamoto earthquake sequence, Tsujun Bridge in Yamato Town, Kamimashiki County, Japan, suffered significant failure including a water leakage from stone canals. In this study, in order to evaluate ground motions at the failure site, microtremor measurements and temporary aftershock observation were conducted at the site. The evaluated site amplification and phase effects between the damage site and K-NET Yabe (KMM009) site are very similar. It indicates that the observed ground motions at K-NET Yabe site during the 2016 foreshock and the 2016 main shock can be regarded as the ground motions at the failure site.

Accurate evaluation on the shaking characteristics of an embankment is very important for the rational seismic assessment. In this study, the transfer functions of a road embankment were evaluated based on the in-situ records of seismic observation and microtremor measurement. The both transfer functions were almost similar, indicating the possibility that the shear wave velocity profile of an embankment using the microtremor measurement can be evaluated by the original equation for natural frequency estimation and/or the dynamic linear 2-D FEM analysis.