Probabilistic seismic hazard analysis (PSHA) was implemented and the exceedance frequency of peak ground acceleration (PGA) on the ground surface for aftershocks over one year following the 2011 Tohoku earthquake (M_w9.0) was estimated. The magnitude-frequency relationship for aftershocks was estimated using two models, in view of the validity assessment and prediction. The first used the aftershock catalog of the Tohoku earthquake (Model 1) and the second used the catalog of the aftershocks accompanying the mainshocks occurring before the Tohoku earthquake with a smaller magnitude up to 8 compared to the Tohoku earthquake (Model 2). Exceedance frequency was estimated at four stations of the Strong-motion Seismograph Networks, operated by the National Research Institute for Earth Science and Disaster Resilience of Japan, along the coastline of the Pacific Ocean in the Tohoku and Kanto districts, Japan. The estimated frequency exceedances for the PGA based on the two models are mostly equivalent for all stations except for one station which is far from an aftershock zone. Moreover, exceedance frequencies at the four stations were directly compared to those estimated based on the acceleration seismograms recorded at these stations. The results show that predicted exceedance frequencies for both models are nearly consistent with those observed for a PGA less than 200 cm/s² at all stations. However, the former are not consistent with the latter for a PGA significantly greater than 200 cm/s² at some stations. Finally, exceedance frequencies considering both the mainshock and aftershocks were estimated based on the sequence-based probabilistic seismic hazard model proposed by Iervolino et al. (2014). Here, Model 2 was used for the magnitude-frequency relationship of the aftershocks. The results show that the exceedance frequency from the aftershocks is greater than 20% of the total exceedance frequency (which was the combined exceedance frequency from the mainshock and aftershocks) within a PGA range from approximately 300 to 1500 cm/s² (at which buildings may be damaged) for the three stations. Our results suggest that it is important to consider the effects of aftershocks when a PSHA is conducted for a site with the potential of a mega earthquake occurrence, such as the Tohoku earthquake, in terms of earthquake disaster prevention following an earthquake disaster.

In this study, we developed a spatial interpolation method for the high accuracy map of the seismic intensity measure to evaluate structural damage to houses and buildings for crustal earthquakes in Japan using Simple Kriging method. The seismic intensity is calculated using 1-1.5 second 20% damped response spectra, which is the equivalent period and damping when houses and buildings are damaged severely. The developed method can consider not only the effect of shallow soft soils but also the effect of deep sedimentary layers. We confirmed that the distribution of the seismic intensity measure can be estimated accurately using the developed method.

In 2017, the Japanese government started a new operation system that issues earthquake forecasting information for the Nankai trough earthquake. However, the information will remain ambiguous in terms of where, when, and how large destructive earthquake occurs. Although this ambiguity is inevitable due to the uncertainties inherent to the information, the overall impacts to disaster mitigation are not well understood. We develop a new scenario method based on the idea of scenario planning approach to identify the effects and potential problems in utilizing the information for disaster mitigation. In this study, we constructed a scenario for a “slowslip” case where anomalous crustal deformation is observed but cannot be determined as precursor of the Nankai trough earthquake. The scenario described how and when the Nankai trough earthquake information is issued on the onset and development of anomalous “slowslip” crustal deformation and how the information is explained and disseminated through TV news media. Discussions among seismologists, journalists, and municipal officials for that scenario identified several unrecognized problems which could be caused by inconsistent responses between government and news media, demonstrating that the new method could be used in finding unexpected side effects associated with the Nankai trough earthquake information.

Earthquake disaster prevention systems using early warning seismometers have been introduced to ensure safety during an earthquake for high speed rails in Japan and have quickly suspended the trains during past earthquakes. An early warning seismometer estimates the epicentral location and magnitude just after detecting the arrival of P-wave when an earthquake occurs. Because the seismometers utilize weak initial P-wave information to estimate seismic parameters, it is expected that the installation environment of the seismometers affects the estimation accuracy, but this effect is not clear. In this study, in order to improve the estimation accuracy of the epicentral location by the seismometers, we evaluated the relationship between the characteristics of seismic observation sites and the estimation error of epicentral locations. Then we suggested the selecting method of the seismic observation site and confirmed its effect.

The purpose of this study is to generate nationwide maps for liquefaction susceptibility. Random forest, which is one of the machine learning methods, was selected to solve a binary classification of liquefaction as opposed to non-liquefaction. Our dataset consisted of 16 variables related to soil density, soil saturation and earthquake ground motion. Furthermore, the dataset has a highly imbalance problem of the classes, because the number of approximately 18,000 cells are liquefaction grid cells while the number of approximately 115 million cells are non-liquefaction grid cells. To solve the imbalance problem, we proposed an ensemble method combining under-sampling. As a result, the proposed method achieved an overall accuracy score of 95.1%, a recall score of 91.4% and a precision score of 7.6% with the imbalanced data. Finally, a parametric study based on seismic intensity was conducted to create liquefaction susceptibility maps.

Due to the 2011 off the Pacific coast of Tohoku Earthquake, liquefaction occurred over a wide area and caused enormous damage. After that, a maximum seismic intensity of 7 was observed during the 2016 Kumamoto Earthquake and the 2018 Hokkaido Eastern Iburi Earthquake, and liquefaction occurred over a wide area. In this study, we selected earthquakes that can count liquefaction points using detailed aerial imagery. The selected earthquakes were the 2011 off the Pacific coast of Tohoku Earthquake, the 2016 Kumamoto Earthquake, the 2016 Tottori Prefecture Central Earthquake, the 2018 Osaka Northern Earthquake, and the 2018 Hokkaido Eastern Iburi Earthquake. Based on the information on the liquefaction occurrence points of these earthquakes, the liquefaction occurrence ratio based on the 250 m mesh unit, and the liquefaction area ratio considering the liquefaction area within the 250 m mesh were estimated. Furthermore, we proposed a formula for estimating the liquefaction risk ratio by the product of the liquefaction occurrence ratio and the liquefaction area ratio.

We believed that in order to accurately evaluate the level of damage to buildings in a large earthquake, there was a need to understand the input motion of each building, so we developed a damage meter for residential buildings. This paper involves an experiment using these damage meters set up in an area of Machida, Tokyo, with sloping ground and a lot of wooden residential buildings close together, in order to test the input motion. Specifically, using the input motion indices of seismic intensity, peak acceleration, and input acceleration trajectory, we tested the effects of a building's elevation, its direction, and the surface ground. As a result, we found that even if buildings were right next to each other, the values and forms of seismic intensity, peak acceleration, and input acceleration trajectory were different for each building and each earthquake, and we discovered their characteristics.

We need a method that can predict strong ground motions with sufficient accuracy at any target sites. So far, however, our knowledge about the characteristics of source, path, and site factors of the observed strong motions has not been fully utilized. First, we performed the analysis to investigate the properties of the factors mentioned above based on a generalized inversion technique (GIT) on Fourier spectra of strong motion networks deployed in Japan, then we tried to model not only spectral amplitude but also phase in spectra. Next, we constructed a procedure for predicting the strong motions considering both the spectral difference between the whole duration of motion and the S-wave portion and the effects of soil nonlinearity. Finally, we confirmed the proposed method worked well for the largest aftershock (Mw7.8) of the 2011 Tohoku earthquake with a special consideration of its smaller stress drop as a regional source characteristic.