Large tanks for storing petroleum are placed in various locations in Japan. The natural periods of liquid sloshing of these tanks are calculated from their diameters and contained liquid heights. The natural periods are approximately between 5 to 10 seconds and hence, petroleum tanks are affected by long-period seismic ground motion. Thus the sloshing phenomenon, oscillation of the contained liquid, may occur and sometimes fire will be associated. Considering this background, this study focuses on the sloshing phenomenon and the sloshing heights in the 2011 Tohoku-Oki earthquake were calculated to assess the fire risk in the Keiyo petrochemical industrial complex. The long-period ground motion due to the scenario Nankai-Trough earthquake was also employed for sloshing analysis at this site. From the calculation of sloshing heights and overflow amount of oil under various conditions, the method to prevent oil overflow was suggested for future earthquakes.

In this study, a new calculation procedure of near-source displacement waveform in frequency domain is proposed. This procedure evaluates far-field term, intermediate-field term and near-field term using ω^-2 model as the moment rate function. It has been thought that this kind of analyses were very difficult, because so-called permanent displacement included in the intermediate-field term and near-field term causes drift problem. The proposed procedure resolves this drift problem, and acceleration, velocity and displacement time histories remain fully compatible and have causality.

A procedure is presented for evaluating the liquefaction-induced residual displacement by static analyses. This procedure is composed of four analyses. At first, existing stress state is evaluated by, so called, initial stress analysis, which is a sequential analysis following the construction process. Then liquefaction analysis is made to obtain the liquefaction resistance factor F_L. This analysis can be a simplified procedure or total stress seismic response analysis. Following it, deformation analysis is made in which stress-strain curve after the liquefaction is used. Finally, volume change caused by excess porewater pressure dissipation is evaluated by the simplified consolidation analysis. In order to complete this procedure, laboratory tests are carried out to obtain the stress-strain curve after the liquefaction. It is expressed by a concave type bi-linear model and the two stiffnesses are expressed as a function with respect to F_L and liquefaction strength ratio. It is also found that behavior of non-liquefied soil is important through numerical analysis and a new stress-strain model based on elastic- perfectly plastic model with Mohr-Coulomb/Dracker Prager type failure surface is introduced.

This paper shows a hazard to lead to a flood risk caused by ground deformation due to the Uemachi fault zone earthquake in the east area of the Osaka plain. First, the ground is assumed a homogeneous semi-infinite elastic medium, the fault parameters are set based on results of study and research of the past, and vertical displacement of earth's surface caused by dislocation of a rectangular fault in this medium is analyzed by numerical simulation based on the Okada's formula (the Okada's Dislocation Model). Next, it is grasped that a trend of ground deformation in the east area of the Osaka plain due to one event in the Uemachi fault zone. And, we paid attention to sedimentation situation of Ma12 and Ma10 layer which is diluvial clay formation make up the Osaka plain, and analyzed this situation and numerical simulation results with deformation based on the estimated number of activity times after formation of these layers started. As a result, about tilting rate of strata in these periods, we found that numerical simulation results are correlated well with analysis results of geologic strata by geotechnical data in the east area of the Osaka plain. Moreover, it makes clear that wide area ground deformation in this area due to one event in the Uemachi fault zone can become a hazard to lead to a flood risk because this ground deformation will cause reverse gradient of some rivers in this area.

To analyze why a lot of buildings collapsed in Mashiki town by the 2016 Kumamoto earthquake, we compared the distribution of collapsed buildings with the construction years, the result of microtremor measurements of the ground, and the earthquake ground motions. The collapsed buildings and the construction years were judged from some aerial photographs. This study reveals that the damaged area shapes like a band, and the band caused by the foreshock is different from the band caused by the main shock. The collapsed building ratio is closely related to the maximum value of the pseudo velocity response spectrum calculated by the earthquake ground motion. The severest area includes many collapsed wooden buildings built after Order for Enforcement of the Building Standard Act amended in 1981. In the same area, the ratios of the collapsed buildings built before 1982 and after 1982 are almost same.

About 44% of the people who evacuated from the giant tsunami of the East Japan Great Earthquake returned home before starting evacuation or dropped in before they reached a safe place. The author used interview data of survivors collected by the City Bureau, the Ministry of Land, Infrastructure, Transport and Tourism; and the Center for Space Information Science of the University of Tokyo, and analysed the purposes and consequences of returning home or dropping in by classifying the data into the rias coast area and the plain coast area, and into foot evacuation and automobile evacuation.

Notable among many outcomes were:

(1) When evacuees dropped in during evacuation, the time required to escape from tsunami inundation and reach a safe zone, became 3.2 times for foot evacuation in the plain coast area, and became 3.6 times for automobile evacuation in the rias coast area.

(2) Safety confirmation and rescue of family, relative, and acquaintance were the most common purposes of returning home and dropping in. For dropping in by automobile, these accounted for 58% in the rias coast area and 64% in the plain coast area.

(3) When the required times to reach a safe zone by foot evacuation and automobile evacuation were compared, the results were different according to the following combinations: evacuation in the rias coast or plain coast, evacuation after returning home or coming from home, and evacuation with or without dropping in. But, the data of number weighted averages of the required times for all combinations were almost the same for foot evacuation and automobile evacuation.

We verify the validity of the depth of the engineering base surface for seismic response analysis of public infrastructures in the Osaka plain. The main point of this paper is as follows. First, At an alluvial plain we set the engineering base surface on a layer that is S-wave velocity equivalent about 300m/s that is the ground where it is admitted that it is firm in designing public infrastructures. Second, we select some verification points that is the representative of alluvial ground of the Osaka plain that is formed deposition of successive alternated of gravel layer (Vs≥300m/s) and clay layer (Vs≤300m/s) under the alluvial layer. And each site is applied one-dimension non-linear seismic response analysis, we verify maximum acceleration and shear strain at ground surface by seismic wave input to each of the upper of the Pleistocene gravel layer (Dg2-layer) shown as the engineering base surface in previous studies and lower of clay layer (Vs≤300m/s) under Dg2-layer. Moreover we evaluate the engineering base surface proposed in previous studies in order to examine the seismic performance of public infrastructures.

We carried out damage investigation around the seismic stations where high JMA seismic intensity scales were recorded in the 2014 Northern Nagano Earthquake. We found some minor damage such as small damage to the wall, but heavy damage to buildings was not found around all the seismic stations. We investigated the correspondence of strong ground motions with damage to buildings. Very short period below 0.5 sec was dominated in most strong motions and the 1-1.5 sec response which has close relationship with heavy damage to buildings was small, therefore, heavy damage to buildings was not found in spite of high JMA seismic intensity scale.

The Q value and short-period spectral level of the 2016 Kumamoto earthquake and its aftershocks are evaluated using spectral inversion method. The estimated short-period spectral levels A are 1.3∼1.8 times larger than expected by Satoh's empirical relation for strike-slip earhquakes. But, the estimated A is similar to that the Dan et al's empirical relation. The estimated Q value is modeled as 79f^0.78 in a frequency range of 0.5-10.0Hz.