Journal of Japan Association for Earthquake Engineering Vol. 17(2017) No. 2

This study is aiming to construct source model for simulating long-period motions (2-10 seconds) that cause damage of long-period structures (e.g. Skyscrapers, and large petroleum storage tanks) for mega-thrust subduction earthquakes. The large slip area caused catastrophic tsunamis was located on the shallow area near the Japan trench during the 2011 Tohoku earthquake. On the other hand, the strong motion generation area (SMGA) generating the strong motions were located west of the hypocenter and along the down-dip edge of the source fault. We clarified that ground motions from the SMGAs have the omega-squared spectral characteristics including long-periods from 2 to 10 s of engineering interest, analyzing the observed records identified as wavepackets from the SMGAs Then, we propose a method of evaluating the long-period ground motions from the SMGAs using an appropriate subfault size and slip velocity time functions to be able to follow the omega-square scaling model.

The Mikawa earthquake (M=6.8) occurred on January 13th in 1945, which was about one month after the Tonankai earthquake (M=7.9) and was in the midst of the Second World War. The Nk value is defined as the number of the totally collapsed wooden houses divided by the number of the death tool to discuss the characteristics of damages from the Mikawa earthquake. The average of Nk values is estimated to be about 10 for the 10 disaster inland earthquakes from 1891 to 1995 in Japan. The Nk value from the Mikawa earthquake shows 3.1, which is meaningfully small compared with other inland events. In this paper, causes of the small Nk value of the Mikawa earthquake are discussed. An influence of the evacuation in group for school children and teachers under the war can be estimated to be 0.4. Without this influence, Nk becomes 3.5, which is still small compared with the average. Influences of the occurrence shortly after the Tonankai earthquake and of the occurrence at midnight in winter were also considered to reduce the Nk value, though they cannot be quantified. On the other hand, it is found that Nk values show smaller tendency in the communities including the trace of the earthquake fault than those in the other municipalities. We examine the detail locations of totally collapsed wooden houses and death toll in Kota Village and the Katahara Town, where the surface fault breaks clearly appeared on the Fukozu fault. It was found that the Ichiba area of Koda Village and the northern Katahara Town in the vicinity of the Fukozu Fault show especially small, that are Nk=1.2 and 1.1, respectively. In these areas, large vertical displacements with more than 100cm can be found at many points along the Fukozu fault. It may be concluded that one of the highly probable causes of small Nk value near the earthquake fault is the influence of the occurrence of many cracks and inclination of ground accompanied by the surface fault break on addition to the extremely strong ground motion. In these situations, houses may be suddenly collapsed and people cannot avoid the risk of death easily.

In order to clarify the spatial variation of the ground motions, we carried out verification tests for evaluation of the spatial variation of the ground motions using high-density horizontal array observation records of 10 observation points in the Quaternary ground having a substantially horizontal geological structure. As part of the verification tests, we performed Monte Carlo simulations using two-dimensional heterogeneous ground models estimated from the in-situ ground data, and attempted to estimate the spatial variation of the ground motions. As a result, based on the ground heterogeneity obtained from the in-situ ground survey, it is possible to estimate the spatial variation characteristics (spatial correlation coefficient, variation) of the ground motions in the local ground with a range from several meters to 200 meters.

This study evaluates the number of demolished wooden houses after the scenario Tokyo Metropolitan earthquake. The building dataset compiled by Kobe City Government after the 1995 Kobe earthquake was employed to achieve the objective. A method to estimate the number of demolished wooden houses due to ground shaking is developed considering the spatial distribution of the JMA (Japan Meteorological Agency) seismic intensity. The result after the 1995 Kobe earthquake was compared with that after the 2004 Mid Niigata earthquake, and the difference between the two events was not significant. Lastly, the method was applied for the scenario Tokyo Metropolitan earthquake to estimate the number of demolished wooden houses.

The seismic damage estimation using the empirical attenuation relationship is conducted for evaluating effects of spatial correlation between intra-event residuals. By approximating the correlation matrix in a sparse matrix, we can generate intra-event residuals with considering a spatial correlation more efficiently and apply this method to a great earthquake. We estimate seismic damages on the Nankai Trough Earthquake and the Sagami Trough Earthquake. By comparison with seismic damages calculated with spatially correlated residuals and spatially uncorrelated residuals or perfect correlated residuals, we found the frequency distributions of damaged buildings are different. The result indicates our method takes into account diversity and uncertainty of earthquake in the seismic damage estimation simply.

In this study, we develop two methods for the inference of rotation vector on ground surface, two rocking rotations and a single torsional rotation. The first, termed nth-order elastic method, is based on the elasticity of the ground surface, and the rotation vector is constructed from the first derivative of ground motions which are approximated by nth-order Taylor expansion. Meanwhile, the second, termed rigid method, is based on the rigidity of ground surface and the rotation vectors are directly obtained by the least square method which minimize the sum of the squared difference motions recorded and predicted by the equation of the rigid motions. Furthermore, the second is divided into two methods. The one, multi-site rigid method, uses the differential motions at multi-sites, while the other, single-site rigid method, uses the differential motions at a single site. Also, we show that the 1st-order elastic method is the same as the multi-site rigid method. Appling the nth-order elastic method to microtremor recordings acquired with a small-size dense array, we successfully infer the rotation vector and find the rigid zone with a radius of 5m for the torsional motions. We furthermore obtain the two findings. The first is that the rotation vector can be inferred with a simpler array of a small size and a fewer observation sites. The second is that the root mean squared (RMS) amplitudes of the torsional motions inferred by the single site rigid method are approximately the same as the RMS amplitudes inferred by the 1st-order elastic method in a zone close to the reference site.

It is necessary to deal with a variety of uncertainties to affect the prediction of tsunamis in the probabilistic tsunami hazard assessment method. The rupture process parameter in a tsunami source model is one of them. In order to grasp influence of the uncertainties regarding the rupture process parameter in the evaluation of tsunami water level, we conducted tsunami simulation with a lot of tsunami source models which changed the location of rupture start point and rupture velocity assuming the concentric circular rupture propagation, estimated tsunami water levels, and analyzed its influence quantitatively. In addition, based on the analysis, we developed a simple method to deal with the uncertainties regarding the rupture process parameter in the probabilistic tsunami hazard assessment.

A 98-dimensional feature vector has been proposed to represent temporal characteristics of a waveform of ground motion acceleration on the basis of its Husid plot. Dissimilarity of a pair of waveforms can be quantified in terms of Euclidian distance between corresponding feature vectors. Then a method has been proposed to cluster and classify temporal characteristics of waveforms by use of hierarchical cluster analysis. Illustrative examples using acceleration records observed in the 2011 Off the Pacific Coast of Tohoku Earthquake, Japan are shown. The distribution map of the clustering and classification show that the temporal characteristics affected by the distance from the source region and multiple-source process can be appropriately classified. Multidimensional scaling is also applied to characterize the result of clustering and classification. It is shown that hierarchical cluster analysis is capable of reflecting temporal characteristics better than significant duration.

The damage caused by the 2011 Tohoku earthquake (Mw = 9.0) to water transmission and distribution pipelines in Japan is investigated using the database on pipeline repairs provided by the Ministry of Health, Labour and Welfare. The numbers of repair are summarized according to areas of water supply corporations and the presence or absence of liquefaction during the Tohoku earthquake. The locations of the pipeline repairs are discussed with respect to the earthquake intensity and geomorphologic conditions. The following findings were obtained: (1) Approximately 25% of the pipeline repairs, with a pipe diameter of more than 50 mm, occurred at liquefied areas and in the pipelines installed in the liquefiable geomorphological units, such as artificial fills. However, 75 % of the repairs occurred outside the liquefied areas and in the pipelines installed in the unliquefiable higher lands, such as the hill and terrace areas comprising hard soils; (2) There was a distinct difference between the Tohoku and Kanto regions in terms of the numbers of repair in the liquefied areas, and the pipelines in the Kanto region were strongly influenced by liquefaction; (3) The number of repairs were counted according to the earthquake intensity and depending on the presence or absence of liquefaction. Pipe repairs were required in the areas where the I_JMA exceeded 5-upper, the PGAs exceeded 150 cm/s², or where the PGVs exceeded 20 cm/s in the liquefied areas; (4) However, the weaker the earthquake intensity, the greater will be the numbers of repair in the liquefied areas. This implies that the pipelines were damaged not by the shaking of the ground, but by the liquefaction that occurred in the areas where the earthquake intensity was not very strong; (5) Counting the numbers of repair with respect to the pipe material and coupling in the area where severe liquefaction occurred, no tendency was found to indicate that the specific material or coupling was selectively damaged; (6) The damage ratio relative to the pipe materials varied considerably according to the area of the water-supply corporation where severe liquefaction occurred; (7) The likelihood of liquefaction-induced damage was evaluated at 6 sites in the liquefied area where many pipeline repairs occurred 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. From the result of the analysis, it was found that all of the sites were evaluated as “the highest likelihood of liquefaction-induced damage”.

Mortality rates caused by tsunamis vary from community to community, depending on geographical and social features peculiar to each. If the relation between mortality rate on the one hand and geographical and social features on the other can be quantitatively formulated, it can be a means to concretely evaluate the community's vulnerability with regard to evacuation (hereafter, evacuation vulnerability) and to adopt measures that effectively reduce loss of human life.

Therefore, the authors proposed to apply HVI (Human Vulnerability Index), defined as the rate of mortality divided by the rate of incidence of washed-out buildings, to evaluate the evacuation vulnerability of a municipality. Then, using reliable public databases, the authors evaluated the HVIs of the twenty municipalities that were heavily damaged by the tsunami of the 2011 Great East Japan Earthquake. And, they applied a multiple-regression analysis using the following four factors as explanatory variables and extracted a reliable formula (R=0.908), which enables us to analyze the evacuation vulnerability more rigorously. The formula also enables us to evaluate the HVI of a municipality before being suffered with a large tsunami.

1) Allowance: Tsunami arrival time after earthquake divided by length of evacuation route;

The increase of allowance effectively lowered HVI.

2) Preparedness: Rate of persons who always prepared emergency bag;

Disaster education such as letting people prepare emergency bags was effective in lowering HVI.

3) Road serviceability: Rate of car evacuees multiplied by car speed;

The higher the road serviceability, the lower the HVI.

4) Warning intensity and cognition: Forecasted tsunami heights multiplied by people’s cognition rate;

The forecast tsunami heights initially broadcast to the municipalities' areas was sensitive to HVI.

The river embankment of Eaigawa river system in Miyagi prefecture has been accumulated the damages by the huge earthquakes, such as the 1962 Northern Miyagi Earthquake, the 1978 off Miyagi Earthquake, and the 2011 off the Pacific coast of Tohoku Earthquake. In this report, the characteristics of vulnerable places and the effects on the earthquake by the countermeasure for leakage and scour were evaluated by comparing the damage cases of the Eaigawa river embankment which were collected and organized. The results of the analysis show that the seismic damage is prone to increase as the time is shorter after the completion of river improvement work because the intensity of expression for effect of aging cannot be expected. Moreover, it was found that the seismic damage is likely to reduce in the construction place of the cutoff sheet-pile and concrete block revetment of leakage and scour countermeasures. In conclusion, the effect on the damage reduction by the leakage and scour countermeasures are important in finding for the efficient implementation of the river improvement project due to its long linear structures, but the seismic countermeasure is quite difficult.

Serious damage was caused to Yokoe Bridge in Yatsushiro City during the main shock of the 2016 Kumamoto earthquake. In this report, seismic waveforms during the 2016 main shock in the bridge were estimated using the site-effect substitution method. The site effect at the bridge was evaluated based on temporary seismic observations. Difference of the site effects between the bridge site and the present observation stations was also indicated.

Residential landslides in Miami-Aso Village, which are located near the epicenter of the main shock of the 2016 Kumamoto Earthquake were occurred due to the strong motion. In this study, ground investigations were conducted at the landslide sites in order to evaluate the empirical site effects. Seismic waveforms at the sites of interest 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.