In order to investigate causes of cut-off frequency f_max, that is important for strong ground motion prediction, we conducted deep seismic motion observations at depths of 300 m (Vs = 2.8 km/s) and 1000 m (Vs = 3.2 km/s) of hard rock site. As a result, we showed that high-cut filter based on f_max may correspond to an attenuation characteristic, so-called "bedrock characteristic", in a deep bedrock layer (generally corresponds to the 3.6 km/s layer) deeper than the seismic bedrock which had not been considered before.

This paper presents the analytical results of water pressure lowering of water supply network at the time of an earthquake. Focusing on the phenomenon occurred in Kawagoe City at the time of the Tohoku Great Earthquake, two types of analysis are carried out, one is the sloshing analysis of water storage tank, and the other is the vibration analysis of water in pipe. The change of water pressure of pipe analysis is much larger than the case of tank sloshing. Therefor the compression behavior of water in pipe is influenced on water pressure lowering phenomenon.

In the ‘Recipe’ for predicting strong ground motions from crustal earthquakes proposed by the Headquarters for Earthquake Research Promotion, Japan (2016), the seismic moment is evaluated from the seismic fault area, excluding the fault area located in the surface layer, even when the fault rupture reaches the surface, and this leads to the underestimation of the seismic moment. Hence, in this paper, we examined the effects of two different ways of counting fault areas, that are the seismic fault area and the entire ruptured fault area, on the evaluated fault parameters. Then, we presented three fault models with the fault length of 30 km, 50 km, and 200 km, and we predicted strong ground motions by the stochastic Green's function method. We found that the effects of the different ways of counting fault areas on the strong ground motions from the faults 30 km and 50 km long were small for peak ground accelerations and comparatively large for peak ground velocities and response spectra in the longer period range. On the other hand, the effects on the strong ground motions from the fault 200 km long were little for peak ground accelerations and peak ground velocities while the effects were a little for response spectra in the period longer than 3 seconds.

This paper examines how the internationally-deployed search and rescue team of Japan (Japan Disaster Relief Team: JDR) strengthened its capacity through the IEC (INSARAG External Classification) in 2010 and IER (INSARAG External Reclassification) in 2015, the classification system conducted by INSARAG (International Search and Rescue Advisory Group). This paper concludes that the IEC/R contributed to the strengthening of the team's capability such as change of team's composition and introduction of the new techniques, through learning from other countries, which Japan did not possess. Furthermore, the effect of the IEC/R was not limited to the JDR team, as it also contributed to the improvement of the domestic USAR activities in Japan.

This paper presents a new method of generating the time-history waveforms which are compatible with multiple-damping design spectra. To assess the structural integrity of buildings and reactor buildings, time-history response analyses are conducted using artificial seismic waves which are matched with the target design spectra. The damping factors of artificial seismic waves are defined as approximately from 0.01 to 0.05. These damping factors differ from damping factors applied to seismically isolated buildings because the seismically isolated buildings use the damping forces corresponding to the damping factors of approximately 0.2 to 0.4. The time-history waveforms which are matched with the design target spectrum with damping factor of 0.05 can lead to large variations in the seismic responses for the seismically isolated buildings. The iteration schemes to suppress the large variations in the seismic responses have been developed by employing the correction waveforms. However, these iteration schemes which are capable of being applied to structures with high damping factors, such as the seismically isolated buildings, have not been developed yet. This is because the frequency intervals of the correction waveforms significantly affect the convergence to the target design spectra with the increasing damping factor. Therefore, this paper presents a new type of method focused on the frequency intervals of the correction waveforms and demonstrates the effectiveness of this method by generating the time-history waveforms which are compatible with multiple-damping design spectra including the high-damping factors.

Shitoki Dam is a rockfill dam of 83.5 meters in height located in Fukushima Prefecture, Japan. It suffered severe earthquake motions during the 2011 off the Pacific coast of Tohoku earthquake and the 2011 Fukushima-ken Hamadori earthquake. The long-term observation records at the dam site provide valuable information about the wave propagation at a large strain level in and around the dam body. We conducted Normalized Input-Output Minimization (NIOM) method for the observation records and examined the propagation time of seismic waves in the dam body. The results revealed that (1) the S-wave propagation time increased from 0.153 s to about 0.23 s during the 2011 off the Pacific coast of Tohoku earthquake due to nonlinear behavior of the dam body; (2) this propagation time corresponds to the 53% reduction of the shear moduli of the dam body at the strain level of 2 to 3 ×10^-4; (3) a similar increase and decrease in the S-wave propagation time were observed during the Fukushima-ken Hamadori earthquake; and (4) the damping ratio in the dam body also increased due to the principal motions of the mainshocks and decreased after them in the similar manner as seen in the results of the propagation time.

During the M 7.3 Kumamoto earthquake of April 16, 2016, the long-period ground motion was observed in Tokyo metropolitan area about 900 km away from the epicenter. It is important to understand the characteristics of the incident seismic motion to the Kanto basin and the response of the basin to the incident for evaluation of the earthquake ground motion in the metropolitan area during a large earthquake in the western part of Japan. First, in order to confirm the incident wave, we examined the velocity traces of F-NET stations from the vicinity of the epicenter to around the Kanto region. A remarkable wave packet with the dominant period of about 10 seconds and with the duration time of about 60 seconds was recognized in the transverse component, and it propagated from the epicentral region to the Kanto basin with the apparent velocity of about 3.3 km/s. This wave packet showed dispersion characteristics and was considered to be Love wave. In addition, the waveforms in the Kanto Mountains area were similar to the waveforms in the western edge of the Kanto basin, and the wave packet was considered to be an incident wave to the basin. Next, we examined the waveform features in the Kanto basin. The wave packet was amplified and the duration time of wave packet was extended in the basin area. However, the dominant period of the seismic motion was about 10 seconds at any observation points. It suggested that the influence of the incident wave to the motions in the basin was so large in this event case. The duration of the wave packet increased with the propagation in the basin and tended to be longer on the east side than the west side of the Tokyo bay. From the multiple filter analysis of the velocity waveform and the semblance analysis for several station sets, it was confirmed that the dispersion of the seismic waves influenced the extension of the duration, and that the wave packets were more dispersed at the eastern observation points. Although the seismic ground motion of the north-south component was dominant in the west side of the Tokyo bay, the large wave packets of the east-west component was recognized in the latter part of the wave traces in the east side. Examining the velocity locus showed that the dominant direction of vibration changed with the lapse of time. In addition, the semblance analysis showed that the direction of wave propagation changed with lapse of time. Those results suggested that bending propagation path of surface waves due to the irregularity of the basin also influenced the generation of later arrivals.