For a wave flume with a piston-type wave maker, a method of generating concentrated waves of arbitrary waveform at arbitrary locations in a channel is proposed based on the theory of linear systems using frequency response function and impulse response function from linear wavemaker theory. The results are compared with those of CADMAS-SURF/3D, a numerical wave tank, and show that the proposed method can be used in experiments, although the waveform is affected by nonlinearity. The present method can be applied not only to physical experiments but also to the wave generation of arbitrary waveforms in numerical wave tanks, which have recently become widely used in practical design.

 In the evolution of nonlinear wave trains, four-wave interaction plays an important role in the wave envelope dynamics, which is related to the modulational instability and potentially leads to the occurrence of freak wave. This research conducts a Monte Carlo simulation based on two-dimensional modified NLS equation considering spatial inhomogeneity, and discusses the evolution of weakly nonlinear wave trains in random directional sea states considering the water depth change from the offshore to the onshore. The numerical result concentrates on the influence from the directional dispersion effect on the fourth-order cumulant in the evolution of the wave field in different bottom topography.

 It is necessary to estimate the wave deformations on the reef when planning wave overtopping counter-measures such as seawalls on the reef coast. In this study, the non-hydrostatic wave model SWASH is used to calculate wave deformations on a reef, and its applicability is examined through comparison with previous models. Both quasi-2D and quasi-3D calculations showed good agreement with the results of model experiments and field observations. In particular, it is shown that SWASH is applicable to the calculation of wave deformations on reefs for all incident wave conditions in the case of a low number of vertical layers. The roughness coefficients were shown to affect the wave height on the reef. The limitation on SWASH espetially on inverse gradient topography and the instability of the calculation was confirmed.

 In this study, the two-dimensional and three-dimensional hybrid model by the solver “shallowInterFoam” that works on the OpenFOAM platform, was verified the relevance of its analysis in comparison with the results of hydraulic model experiments on the flow around the breakwater. Hybrid model calculation results were consistent with the hydraulic model experiment data and three-dimensional model calculation results. Furthermore, the practical improvement was achieved by improving the coupling method between the 2D model and the 3D model, such as enabling the application of unstructured grids.

 While the typhoon Jebi passed through the Osaka Bay, although flooding on coastal areas was fortunately not occurred, the heavy storm surge caused inundation damage mainly in watersides such as harbor areas due to wave overtopping on quays. At a wharf including the quays, the overtopping water mass flows into the land side and it partially returns to the sea on backwash. Moreover, the situation of wave overtopping can depend on both the tide level and the wave condition varying by the passage of typhoon. In this study, a Boussinesq wave model is improved so that it can calculate flooding on wharves due to wave overtopping continuously, responding to the change of wave and tide level with time. On the calculation reproducing the inundation on the east-side of Rokko Island when typhoon Jebi was passing over, while the calculated inundation heights underestimate the trace heights obtained by site investigation, it is confirmed that the calculated inundation area varies reflecting the history of flooding and discharge on wharves.

 In this study, a theoretical frequency condition to create the focusing wave packet in uniform current is proposed and validated through wave-current experiments. We observed the maximum surface elevation monotonically increase with the current velocity and the highest elevation of six times higher than the initial wave amplitude is recorded at the highest current velocity in the present experimental conditions. We also found the temporal variations of the spectrum width during the focusing process and the nonlinear steep wave form with double harmonic component at the focusing location.

 In the water-front area facing to Tokyo Bay in Yokohama city, the inundation disaster was occured while the typhoon Faxai passed in 2019. It was reported that multi-peaked wave overtopping on revetments was occurred while swells invating from outside of the bay was superposed on double-peaked wind waves generated in inside. In this study, the multi-peaked wave and the single-peaked wave are generated to reproduce the inundation in the water-front area by using a Boussinesq wave model, respectively. Small differences are found out in comparison between both calculation results while they slightly underestimate the trace levels observed by field survey. Therefore, it suggests that more suitable estimation of invating swell energy in the multi-peaked wave spectrum is important to reproduce the inundation with better accuracy.

 In this study, the novel LES-Stochastic model describing bubble rise, bubble-interface interactions, bubble cluster formations, coupled with turbulent fluid motion, has been proposed. The foam model, developed on the basis of the analytical solution of bubble-bubble mechanical interactions by Nicolson (1948), reasonably simulated oscillatory bubble behaviors beneath the interface and formations of the bubble clusters. These processes create additional turbulence near the interface and thus enhance surface renewal, which may contribute material transport between air and water.

 The directional spectra expressed by a parametric function based on Goda et al. (1975) are commonly used to design use for multi-directional waves. It is proposed based on the extrapolation of parameters from wind wave observations, and the validity for swell conditions has not been confirmed yet. In this study, the validity of the directional spectrum proposed by Goda et al. (1975) was examined based on field observations. The dependence of the directional spreading on wave age and wave steepness was weak, and the directional spreading kept almost constant value. Compared with previous studies, the extrapolation of parameters from Hasselmann et al. (1980) to the swell region was the most consistent, whereas there were significant differences from Mitsuyasu et al. (1975), especially in the swell. On the other hand, the reference value by Goda et al. (1975) for the directional narrowness parameter generally covered the range of directional spreading observed in this study.

 In order to cope with missing data observed by Doppler-type Directional Wavemeter (DWM) and to maintain stable wave observations, we developed a method to recover surface waves from bottom pressure records observed by pressure gauges equipped on DWM. A neural network was used to estimate the transfer function required for the recovery. For the input data of the neural network, bicoherence was used for the purpose of considering the effects of nonlinearity and multidirectionality of coastal waves. Then, the spectrum of the surface waves were estimated based on the estimated transfer functions, and their accuracy was confirmed by comparing the observed and estimated energies in each frequency band. As a result, it was confirmed that the proposed method can accurately recover the water surface in a wide frequency range including long-period waves.

 To predict extreme storm surges with those probability due to extreme typhoons, this study examined capacity of the probabilistic typhoon models which developed by use of the passed typhoon best track in and around the target points of Tokyo, Ise and Osaka Bay in Japan. Though obtained results by the developed model were unclear on the relationship between passage target ranges and anormal storm surges, anormal storm surges became larger in each bay than those by the models which were developed by using whole typhoon best track data. In case the proposed probabilistic typhoon model takes the limit of central pressure degreasing over an open sea, the model is able to calculate realistic anormal storm surges.

 The risk of storm surge disasters is increasing due to the sea level rise and the increase in typhoon intensity caused by climate change. In order to enhance the evacuation framework against the maximum possible storm surge, the flood prevention law has been amended, and then local governments create a storm surge inundation hazard map. However, there is no estimation of the recurrence period of the maximum storm surge. In this study, an ensemble storm surge simulation is conducted in Osaka Bay using a nonlinear shallow water model with the aid of a global stochastic typhoon model to estimate the return period of storm surges caused by Typhoon Jebi, the largest typhoon ever recorded, and by the assumed maximum storm surge. The storm surge during Typhoon Jebi is simulated with sufficient accuracy. The estimated returen period of the storm surge caused by Typhoon Jebi is 50 to 100 years. The estimated return period of the assumed maximum storm surge is more than 2500 years, indicating that the local governments are assuming a very rare event.

 Superimposed hazards by storm surge, height waves and flood in the coastal area have possibilities of accumulating the damage further than expected. The same is true for simultaneous disaster in the neighboring towns due to a single enormous meteorological disturbance. Recently several definitions of return period for combinatorial occurrence of multivariate extremes are suggested by using the copula of multivariate statistics, but those seem to be very complicated for us to make plans for flood prevention and reduction. This study propose a new simple index, as named by accordance, for joint occurrence of multivariate extreme events, and the statistical simulation, whose procedure is developed newly and adjusted for this purpose, are conducted to reveal the properties of correlations and accordance by employing an asymmetric family of bivariate extremes.

 In maritime construction, there is a high demand for wave prediction to determine whether work is possible or not. In this study, we used ConvNet (Convolutional neural network) for deep learning, and built models to predict the wave in 24 hours at the NOWPHAS observation point using the JMA air pressure data. Since there are only a few thousand data available for training, we thought there would be a shortage of data for deep learning. To solve this problem, we introduced Fine Tuning (FT) into the model, which is a method to train the model efficiently by using the learned model. And we confirmed the effectiveness of the model. A comparison of the prediction accuracy of the models shows that FT is more efficient and accurate than conventional learning methods in deep learning with a limited number of data, demonstrating the effectiveness of FT. Using deep ConvNet, it was shown that the availability of work after 24 hours can be judged with high accuracy even if the training data is only air pressure.

 Studies on the effects of climate change on waves and storm surges have been widely conducted. In comparison with storm surges, there is a lack of information on the future prediction of waves because of the high computational cost. One of the methods to reduce the computational cost is the statistical wave model. Wind speed and pressure are used as predictive variables, but they are not accurate because local equilibrium is not established in the sea area where seasonal changes of swell are dominant. In this study, by introducing the principal component information of the atmospheric field, a highly accurate statistical wave model is developed by adding the atmospheric field in a wide area as an explanatory variable.

 Deep learning technique can be one of promising options for time-effective wave forecasting system. One of shortcomings of these technique however is that the predictive ability of the model is purely based on the training data, and thus the developed system may not be able to represent the phenomena unexperienced in the past. On the other hand, such shortcomings may be useful to detect unexperienced phenomena: such as the influence of the climate change. To answer this hypothesis, this study applied long short-term memory (LSTM) recurrent neural network technique, and developed a model for forecasting of significant wave heights as functions of the recent history of local wave heights and wind speed. The model was trained based on the data observed from the year 2000 to 2007, and the trained model was tested against the data observed after 2008. While the model showed reasonably good predictive skills of the wave height, the model tended to underestimate the extreme wave heights and this underestimation showed increasing trend in the recent years. Similar long-term trend of underestimation was also found in the error of wind speed. Underestimations of wind and wave showed linear relationships, which were consistent with empirical Wilson's formula, meaning that amplification of extreme wave height is mostly due to intensification of the wind speed.

 A screening method for a lot of typhoon scenarios that trigger high wave events is necessary as a preliminary step for detailed simulations using spectral method with large computational load, in order to assess the low frequency catastrophic wave risk. The significance wave method, such as SMB method, is constructed for the case of uniform wind, and it is necessary to devise the definition of representative wind speed and fetch when it is applied to typhoon. In this paper, we developed a fast model SWET, a kind of significant wave method, that can be used for wave estimation. In this study, the accuracy of SWET was verified by comparing it with the spectral method in the open sea near the Kii Channel. The results showed that the accuracy of the estimation decreased when the target area was in shallow water or when the distance between the typhoon track and the target area was large. However, the maximum wave heights can be generally estimated, and the rank correlations are high with the SWAN results. In the current model, the computational time required for wave estimation for one typhoon is about 30 seconds, which indicates the possibility of using the model as a screening method for low-frequency high wave events.

 Typhoon No. 21 in 2017 and Typhoon No. 19 in 2019 passing through Suruga Bay caused severe damages around Shimizu port such as the outflow of harbor cargos to the sea. In this study, the factors of the inundation were analyzed and disaster mitigation measures were examined by an inundation prediction model for the purpose of reducing damages to the port and improving disaster prevention capabilities. Resultantly, it is confirmed that the swell propagating to the port is one of the main factors of inundation in addition to high waves and storm surges caused by typhoon. Numerical simulations of inundation prediction were conducted to evaluate the disaster mitigation measure by using a Boussinesq equation model which can well reproduce the measured sea level in Typhoon No. 19 in 2019. As a result, the seawall on the wharf indicated the reduction effect of the overflow on the wharf.

 In the recent storm surge disasters, the outflow of cargos in harbors and the drifting of ships have been reported. Due to the wind force of typhoon, the behavior of drifting objects during storm surge would be different from that of tsunami. To understand the behavior of drifting objects is an important issue for disaster prevention because global warming is expected to increase the strength of typhoon in the future. In this study, we investigated the behavior of drifting objects around Nagoya Port under the potential maximum typhoon expected in the future. The numerical result shows that the drifting objects are strongly affected by the strong wind of the typhoon. By constructing the geometry model with consideration of building configuration and layout, it is possible to represent the behavior of objects trapped by buildings and drifting between buildings. In addition, for the purpose of use in the disaster prevention measure and education, we created the full 3D computer graphics showing the behavior of drifting objects based on the numerical results.

 We investigated the relationship between the forward speed of a typhoon and wave-induced setup at Hidaka Bay, Wakayama Prefecture, where a remarkable storm surge caused by 2018 typhoon Jebi was observed. The storm surge hindcast was carried out by the ocean-wave coupled model (COAWST), and showed that the maximum storm surge (1.7 m) at the bay was mainly due to wave-induced setup (above 1.2 m). It also maintained the surge height up to 0.5 m for several hours after passing the typhoon, which did not reproduce well by the ocean model alone. Furthermore, we conducted numerical experiments and revealed that when the forward speed exceeded the group velocity of waves, the maximum significant wave height decreased significantly (below -4.0 m) compared to when the forward speed was close to the group velocity. Due to this reduction, the wave-induced setup was being smaller by nearly 1.0 m.

 Strong typhoons can cause storm surges and river floods. A simultaneous occurrence of both along estuaries will cause compound flooding. A numerical model representing the run-up of a storm surge into rivers is necessary for simulating compound floods. In this study, a coupled wave-surge model integrating river inflows was developed to consider the run-up of the storm surge into rivers. The model successfully simulated the storm surge and river water level in Yodo and Ina rivers during Typhoon Jebi (2018). Then, a series of sensitivity experiments on river inflow from the upstream were conducted. As a result, there was no significant difference in water level between coupled and uncoupled models near the river mouth. However, a maximum difference in water level up to 1.2 m at Kema (14 km upstream) was revealed. In addition, the peak time of water levels also changed between coupled and uncoupled models. Therefore, it is necessary to consider both a storm surge run-up and upstream water level to evaluate water levels in a river caused by storm surges and flooding in an estuary.

 It is becoming important to consider the growth of wind waves due to strong typhoon winds in the bay. However, wave growth in the bay is an intermediate phenomenon hardly modeled by either spectral wave model or phase resolving wave model. Therefore, it is necessary to introduce wind stress terms in the phase resolving wave model that doesn’t usually deal with wind wave for inner bays. In this study, we implemented the wave growth term into the phase resolving wave model by considering wave scale momentum exchange between wind and wave interactions. A spectral wave model was used to optimize the parameterization of wave growth in the wave resolving model. The obtained model was then applied to the case of Kansai Airport inundation by the extreme waves of Typhoon Jebi in 2018 to estimate the overtopping volume. Besides, we compared it with the measured values.

 Wave predictions are used in various fields. Marine construction management is greatly influenced by the accuracy of wave forecasting. Wave predictions are made based on the spectral wave model forced by meteorological information, but most are mainly deterministic short-term predictions. On the other hand, it is essential that ensemble wave height prediction for obtaining prediction variability and probabilistic prediction information. In this study, a monthly ensemble wave height forecast was carried out using the statistical wave model. It was confirmed that the results of the ensemble wave prediction generally agree with the observed wave. However, it should be noted that it is underestimated during high waves. In addition, the error is larger when the spread is large, and the tendency was remarkable on the Sea of Japan side.

 In order to improve the accuracy of storm surge prediction in real time, it is necessary to improve the accuracy of typhoon prediction. For this purpose, it is important to properly set the calculation conditions of meteorological models, which have often been determined empirically. In this study, we developed a system using machine learning to automatically suggest combinations of physics options of the WRF to achieve the highest prediction accuracy for newly generated typhoons. As a result, we confirmed the improvement of typhoon forecasting accuracy by using this method, which selects physics options of WRF based on typhoon characteristics such as the position of typhoon occurrence, the maximum diameter of the wind field and so on. In addition, by using the typhoon prediction results as the external force for storm surge estimation, the prediction accuracy of maximum storm surge was improved on average in many locations, and in some locations the accuracy was significantly improved, indicating the effectiveness of our method, although further verification is required.

 In large-scale marine construction works such as offshore wind farm projects, accurate sea state information with a long lead time is desirable. Meteorological organizations provide global wave forecasts, but they have low spatial resolutions, causing problems when used for wave predictions along Japan's coast. A previous study showed that Artificial Neural Network could successfully convert the initial forecast data of Japan Meteorological Agency's Global Wave Model to coastal waves. However, it is not clear whether ANN can well convert the actual GWM forecasts. In this study, we examine the applicability of the wave prediction model, GWM to ANN, by comparing it with coastal waves at Japan's coast.

 The time series prediction of wave height is important for the countermeasure of high wave disaster in the coastal area. Since time series prediction of wave heights over a long recurrence period by dynamical methods is computationally expensive, statistical methods such as neural networks (NN) are considered for prediction. Among the deep learning methods, Long Short Term Memory (LSTM) is considered to be suitable for time series forecasting of wave height because it is good at handling long-term time series data. However, no research has been published so far on time series forecasting of wave height using LSTM. In this study, time series forecasting of wave height is performed using LSTM. The effects of temporal factors such as the input of explanatory variables at multiple times in the past, spatial factors such as the input range of the meteorological field, and the combination of the parameters of the LSTM were varied, and the effects of these factors on the results were compared.

 In this study, numerical investigations using a 3D RANS model were conducted for flow and topography change induced by tsunami in an abrupt expansion area in order to reveal the applicability of the model for the phenomenon. The numerical results of flow simulations clarified that settings of side-wall boundary conditions have a huge effect on the reproducibility of a circulation flow in the abrupt expansion area, and the effect of the differences of turbulence models is small for the reproducibility. Moreover, from the numerical simulations of topography change, it was found that the effect of the non-equilibrium of bed-load transport is small under the targeted experimental conditions, and differences of calculation formulas of the suspended-load transport rate greatly affect the reproduction of the topography change.

 In the present study, the bottom boundary layer characteristics under the 2011 actual tsunami conditions are investigated. The linear long wave theory is utilized to estimate the flow velocity from the measured offshore water level by the GPS buoy stations during the 2011 Great East Japan Earthquake tsunami. Based on this result, the characteristics of the bottom boundary layer such as the boundary layer thickness, flow velocity distribution, and bottom shear force were numerically examined by the k-ω model, which is a typical two-equation model. It was found that the thickness of the tsunami-induced bottom boundary layer is much thinner than the water depth. The bottom boundary layer characteristics are not a steady flow type but rather similar to the wave boundary layer type. Furthermore, we proposed a method for evaluating the tsunami-induced bottom shear stress under an irregular tsunami waveform by using the wave friction coefficient, instead of using the common Manning formula. As a result, the estimated bed shear stress by the new simple method was the most agreement with the k-ω model results among other methods.

 In order to quantify the formation mechanism of seafloor topography on a timescale of millennia in the eastern Seto Inland Sea, the Regional Ocean Modeling System (ROMS) was used to evaluate hydraulic quantities and diagnose suspended sediment transport, coupled with the marine geological discipline. Overall morphodynamics was well captured by the diagnostic model, in which the intensity of bottom shear stress associated with prevailing tidal currents and the divergence of bottom residual currents explain the formations of caldrons and sandbanks. Nevertheless, inconsistency between the background hydrodynamics and the resultant topographies were found in multiple places around tidal straits. We showed that marine geological data was quite useful to fulfil a comprehensive description of, e.g., unexplained zonal differences of sandbank formation relative to Akashi Strait through the considerations of coastal transgression.

 We propose the coRaL method (a method of incorporating Random phase model into Logic tree approach), which is a probabilistic tsunami hazard assessment method with a high objectivity by reforming the evaluation method of earthquake asperity model and the concept of aleatory uncertainty among the probabilistic methods using logic tree approach conventionally used in Japan. The proposed method is applied to the cities located along the Pacific coast where the tsunami risk due to the Nankai Trough megathrust earthquake is assumed. We show that the proposed method can evaluate the tsunami inundation depth distribution for each return period (e.g., 150 years, 500 years, 1000 years, 2500 years), and for each percentile considering the uncertainty of tsunami hazard assessment even for the same return period.

 This study conducts numerical tsunami simulations for future Nankai-Tonankai earthquakes and compares the tsunami amplitudes of different earthquake source catalogs along the coastal area of Japan. One of the catalogs is 11 source models proposed by the Central Disaster Management Council (CDMC), the Cabinet Office of Japan. The other is 300 source models of the stochastic tsunami model, which allows for generating source models randomly. Along the Kochi coast, faced to the open ocean, the mean tsunami heights of the stochastic tsunami model are higher than those of the CDMC models. On the contrary, this tendency is inverted at the coast of Osaka Bay. The difference in the tsunami heights can be explained by the mean slip of the dominant fault region. This result indicates that the acceptable depth variability of the large slip in the earthquake fault model can be a major source of uncertainty in the tsunami height.

 The intraplate earthquake that occurred off the coast of Fukushima Prefecture on November 22, 2016 was a normal fault type earthquake whose fault extended in the northeast-to-southwest direction. The strike parameter of its seismic fault was different from that of the north-to-south direction assumed for this area by the tsunami database prepared, at that time, by the Japan Meteorological Agency for the tsunami warning operation. To examine the effect on the tsunami forecasting of the occurrence of an earthquake different from the assumption, this study analyzes the sensitivity of the tsunami height distribution to fault model parameters including the strike. According to the interpretation from the extent of sensitivity to each parameter obtained based on the case study of the 2016 off Fukushima earthquake, the typical prediction error of the tsunami height owing to the different strike assumption of approximately 45° is equivalent to the difference of tsunami heights owing to the wrong determination of the dip angle of 45° and vertical. Or, it is comparable to the prediction error when the centroid is mistakenly determined as approximately as half of the fault length apart.

 Among the indices for evaluating the reproducibility in numerical simulations of past tsunamis, the geometric standard deviation k of the ratio of the tsunami trace height and the calculated water level was the focus, and the value of k and the changing factors were examined. For tsunamis that occurred around Japan and in the Indian Ocean, collection and analysis of reproduction calculation results of 2D shallowwater models using highly reliable trace records were performed. Results show that the criteria for determining the validity of the analysis model using Aida's indices proposed by the JSCE are appropriate. Also, the reproducibility improved as the number of points with large trace heights were included, and that the upper limit of the ratio between the tsunami trace height and the calculated water level decreased as the calculated water level increased. Furthermore, to obtain the same level of reproducibility as the wide-area analysis results in a tsunami caused by an Mw9 class giant earthquake, it is necessary to target a coastal area that is more than half the length of the tsunami source.

 In this study, in order to improve the movable bed model for tsunamis, the concentration data of suspended sediment under high flow velocity conditions of simulated tsunami was obtained by sediment transport experiments, and the relationship between the equilibrium concentration of suspended sediment and the hydraulic quantities was clarified. In adition, the sediment transport experiments were were reproduced by using the movable bed model for tsunamis with the equation of saturation concentration of suspended sediment. As a result, the equation of saturation concentration of suspended sediment fell within the distribution range of the equilibrium concentration of suspended sediment, and the validity of the equation of saturation concentration of suspended sediment was confirmed by direct comparison with the experimental results. The movable bed model for tsunami with the equation of saturation concentration of suspended sediment reproduced the experimental results of suspended sediment concentration and topography change well.

 The existing estimation formula of tsunami waterborne debris impact force targets the collision force at the local point where the waterborne debris collides with the structure. On the other hand, in a plate-shaped structure such as a seawall, the deformation mode differs depending on the position where the waterborne debris collides, and the collision position is uncertain because the debris is floating on the moving water surface, so the impact force cannot be defined. In this study, for a tsunami seawall, a method to estimate the impact force based on the strain energy generated in the seawall structure was proposed. First, the applicability of the estimation formula under the condition of linear stiffness was validated by the in-air tests, and then a estimation method of the equivalent collision force of a plate-shaped structure based on strain energy was investigated by considering the collision period. In addition, from the results of the hydraulic experiments, it was shown that the impact collision force of a floating vessel can be evaluated by the existing estimation formula by using the above-mentioned evaluation method.

 In recent years, roughness coefficients and drag coefficients that take into account the geometry of structures have been proposed, but their accuracy has been validated using inundation depths and arrival times, and there have been few studies on detailed velocity fields. In this study, the accuracy of the individual Drag Force Model (iDFM), which treats the structures in the computational grid as drag forces, was verified through the idealized numerical experiments and the replication of tsunami inundation experiments. In both cases, the iDFM reproduced the inundation area well regardless of the resolution. The velocity field was not reproduced well in the areas where localized flow was dominant, such as roads and alleys, suggesting the need to improve the model.

 When tsunami reaches a coastal area and intrudes into the water intake facility of the power plant, it may cause damage to the facilities and flooding of the power plant. In this study, we carried out hydraulic experiments and three-dimensional numerical simulations using CADMAS-SURF/3D in order to clarify the tsunami flow process into the water intake facility. In some cases, the flow velocity in the pump room was underestimated in the numerical simulation compared to the experiment. Because the reproducibility of turbulence near the wall was considered as one of causes for the underestimation, we introduced the low-Reynolds-number k-ε model to the numerical model, and confirmed the accuracy improvement of the numerical model. From the results of the hydraulic model experiments and the numerical simulation, it was found that the volume of water intake and the length of the pump room have an influence on the flow process of tsunami intruding into the water intake facility.

 Numerous combinations of disaster reduction facilities should be taken into consideration to evaluate the damages of inundation due to probabilistic tsunami hazard. In this study, we implemented it and investigated how the tsunami inundation probability changes by using the levee fragility model. coRaL method was used to obtain the inundation depth exceedance probability and applied to two different terrains. The results showed that the probability was between the case of overtopping and collapse and the case of overtopping but no breach, which was found to be reasonable, and they support the effectiveness of levee fragility model.

 Kaneto et al. proposed a method to evaluate possible future tsunami caused by submarine landslide. However, superposition with tsunami caused by active faults, which is likely to cause submarine landslides, has not been investigated. This study focuses on the distribution of tsunami arrival times within a radius of 100 km from the Kashiwazaki-Kariwa Nuclear Power Station, proposes a method for determining the combination of tsunamis caused by submarine landslides and active fault. In addition, we evaluated the maximum water level due to the superimposed tsunami by the linear addition method of the tsunami numerical simulation and the method by coupled analysis, and it was confirmed that there was no significant difference in the tendency of the analysis results by either method.

 Experiments and analytical studies have been conducted on tsunamis caused by submarine landslides, but there are few research cases compared to aerial landslides. Therefore, the behavior of tsunamis generated by submarine landslides are not well understood. In this study, in order to investigate the characteristics of the tsunami generated by submarine landslides, tsunami experiments using granules and solid models were carried out using a two-dimensional cross-section channel. In order to understand the tsunami water level in the source area due to submarine landslides, the wave height gauges were installed densely. From the measurement results, the characteristics of the tsunami water level due to the submarine landslide were obtained. Furthermore, the flow field near the landslide generated by the submarine landslide was examined by numerical experiments, and the tsunami generation process was shown.

 Since the event of the 2018 Palu tsunami in Indonesia, the countermeasures for reducing tsunami damage have been considered in Palu city where the destructive landslide tsunami event occurred. The tsunami induced by landslide in the Palu bay could reach the coastline of Palu city within a few minutes. On the other hand, there is a possibility that the tsunami induced a subduction-zone earthquake along the Makassar strait could cause the larger inundation area than that of landslide tsunami, although the tsunami could reach the coastline 20 minutes or more after the earthquake.

 In this study, we investigated an effective countermeasure for complex disaster of landslide tsunami occurring in the Palu Bay and anticipated maximum tsunami induced by a subduction-zone earthquake along the Makassar strait in Palu city based on reconstruction plan in the Sulawesi island. We selected the evacuation facilities based on tsunami hazard map of Palu city and assumed the elevated road along the coastline of Palu city to investigate the effect of tsunami damage reduction. As results of evacuation simulation, the elevated road and appropriate selection of evacuation facilities could reduce the mortality rate of the landslide tsunami. Moreover, the importance of establishment of evacuation facilities which could resist against the tsunami induced by the subduction-zone earthquake was implied by the presented examination.

 In 2011, Japan suffered from the tsunami that was beyond expectations due to the Tohoku Earthquake. Therefore, it is necessary to take countermeasures against tsunamis caused by various factors including landslides. First, 2DH3D hybrid analysis method of Yoneyama et al. was applied to a simulation of the past hydraulic model experiments on tsunamis generated by landslides in order to verify its validity. As a result, this method reproduces experimental results with the same accuracy as 3D method and reduces calculation time significantly, demonstrating its effectiveness for landslide tsunami analysis. Subsequently, a numerical experiment was conducted by the same method for the purpose of selecting landslide parameters that particularly affect the tsunami wave height. As a result, it is considered that the shallower the landslide is, the higher the tsunami is. And, the results shows that the landslide mass volume has a great influence on the tsunami scale.

 Compound physical processes of storm cause floods in coastal urban areas. The present study develops a coupled coastal flood model by considering the sequential physical process of surge overflow, wave overtopping/runup over breakwaters, surface runoff due to precipitation, and backwater from sewer systems. In the present coupled coastal flood model, a surge, wave and tide coupling numerical model (SuWAT) simulates the surge and wave. Integrated Formula of wave Overtopping and Runup Model (IFORM) calculates the wave overtopping discharge/overflow over the breakwater. The Storm Water Management Model (SWMM) estimates the surface runoff and backwater at each grid. The overflow rate, the surface runoff and the backwater are used for boundary conditions in inland flood calculations of SuWAT. The proposed model was applied to a simplified bathymetry to investigate the effect of the complex physical process by conducting a series of numerical experiments. As a result, the flood depth in the coastal urban area increased when all physical processes were considered. Earlier peak precipitation and allowed reverse flow caused the maximum flood depth. The results indicate that the developed model successfully simulates the coastal flood due to the compound factors of the surge overflow, wave overtopping, reverse flow of seawater, and backflow of rainwater through sewer systems.

 Due to increasing risk of inundation in storm surges caused by climate change, it is necessary to consider both wave overtopping and overflow for inundation evaluation. Although some researches were conducted to examine the evaluation method of combined overtopping and overflow, it has not been established and the validation by the hydraulic experiments is necessary. The authors also conducted numerical experiments by OpenFOAM to consider overtopping-overflow discharge and inundation depth. In this study, hydraulic experiments were conducted to verify the accuracy of the numerical experiments and the evaluation method of overtopping-overflow discharge and the inundation level were examined. The validity of the numerical-experimental results was confirmed by means of the hydraulic experiments, and the applicability and problems of the existing formula to estimate overtopping-overflow discharge were clarified. The numerical experiments on the inundation depth in storm surge were also verified, and the equation to estimate the inundation level was updated. It can be also used for the levee with a parapet.

 Several studies have proposed several numerical storm surge models in recent years. For instance, the Integrated Formula of wave Overtopping and Runup Model (IFORM) calculates rates of wave overtopping/runup and surge overflow over breakwaters implemented in SuWAT. The model was introduced so that the accuracy of the storm surge-induced flood prediction can be improved. However, the practical application of the model has been made a few. In this study, we conducted hindcast simulations of the bomb cyclone-induced surge in December 2014 and evaluated calculated inundation depths compared to flood heights surveyed and calculated by the conventional Takayama equation. The results showed that the present model has the same accuracy as the conventional equation when the relative freeboard is small. Also, the present model more accurately simulated the bomb cyclone-induced flood in comparison with the Takayama equation. It was found that the present model mitigates the discontinuity of the wave overtopping/runup and surge overflow over the breakwater.

 Typhoon Faxai in 2019 made landfall in the Kanto region with one of the strongest forces in recorded history. The post-event survey reported a characteristic damage situation. It reported that the severe damage was limited to the area around the port of Yokohama despite the fact that strong winds blew over the entire bay. In this study, we focused on the cause of the coastal damage concentrated in Yokohama Port and the generation mechanism of the external wave force that caused it. It is shown that swell from outside the bay tends to concentrate in the area around Yokohama Port. In addition, when the typhoon hit the area, the swell developed due to strong winds. The wave statistics using wave height and wavelength can clearly characterize the external force of waves off Yokohama Port.

 The sediments within the tsunami inundation flow increases their density. In addition, it is known that the horizontal and vertical forces of the tsunami inundation flow increase accordingly. Thus, it is important to consider sediments when we studied flume experiments of tsunami run-up to river. In this study, we simulated the tsunami run-up to river using two types of sediment, such as fine and coarse sands, and studied a flume experiments on the effect of the tsunami on structure and sediment transport. As a result of this study, we have clarified the following three points. The tsunami inundation flow scours the sediments around the structure. In addition, due to the influence of the structure, almost no sand is deposited behind the structure. The density of tsunami inundation flow is higher under the condition that fine sand is on the sea side and coarse sand is on the river side. The wave force that hits in front of the structure increases when fine sand is on the sea side and coarse sand is on the river side.

 We conducted the disaster surveys of large vessels moored along terminals inside harbor basins in the 2011 Tohoku Earthquake Tsunami by using the analytical method on motions of moored ship due to tsunami wave forces and the seismic vibration. However, the disaster phenomena of mooring line breakage and rendering beyond a winch brake holding load, and the ship’s direct contact with breasting dolphins due to fender breakage has not been surveyed in detail yet. An analytical method on coupling motions of moored ship and dolphins during tsunami has been proposed, and its application has been introduced.

 Tsunami source estimation is important for low-frequency giant tsunami risk assessment in the field of paleo-tsunami. There are many attempts in reconstructing inland inundation conditions using the layer thickness, grain size distribution and distribution of tsunami deposits as constraints so as to estimate the tsunami sources that satisfy the reconstructed inundation conditions. However, sediment source (sea or land) of the tsunami deposits has never been investigated so far. In this study, numerical simulations using sediment transport model were performed using three different tsunami sources model proposed for the 2011 tsunami to investigate whether information of the sediment source of the tsunami deposits can be an additional constraint for the tsunami source estimation. As a result, it was suggested that the source of sediment of the tsunami deposits was very different even if the tsunami source models showed similar inland inundation conditions. Although there are still some problems to be solved, it can be concluded that the information on the sediment source of the tsunami deposits is possibly used as a further constraint in addition to the conventional paleo-tsunami tsunami source estimation method.

 The Ino map, which was completed at the end of the Edo period, contains spatial information such as village, county, temples, shrine, castles, and natural place names, in addition to location information such as coastlines and roads at that time. We thought that the effect of the effectiveness on the temporal and spatial changes of tsunami disaster place names and the risk of natural disasters could be verified by utilizing this old spatial information. In this study, we investigated the existence rate and disappearance time of tsunami disaster place names in the Ino map extracted by spatial information analysis, and evaluated the risks of various natural disasters.