This study investigated relationships between meteorological field patterns and heavy rainfall in western Chugoku and northern Kyushu in Japan using Self-Organizing Map (SOM), and specified meteorological field patterns causing past serious debris flow disaster cases in Hiroshima prefecture during past 40 years. In the next, this study investigated features of meteorological fields relating to the heavy rainfall case in July, 2018 causing the historical serious debris flow disaster in Hiroshima prefecture. The past heavy rainfall cases in Hiroshima prefecture showed strong convective activity characterized commonly by low-level jet and high precipitable water. The most serious disaster case in 2018 was classified into dangerous meteorological field patterns leading to heavy rainfall. The meteorological field patterns of the event were similar to that of the heavy rainfall event occurring in July 24, 2009. The common features between both cases showed the formation of widespread heavy rainfall area, and the occurrence of serious disaster both in Fukuoka and Hiroshima prefecture.

 In this study, we predicted monthly rainfall after two months in the Chao Phraya river in Thailand by using global surface temperature image and convolutional nueral network which is one of the Deep Learning. We evaluated accuracy of two prediction models which was different neuron number. As a result, The model in which the number of neurons was undulated was more higher accuracy prediction. The undulation like as increasing, decreasing and increasing. We divided all of prediction results into every month’s data and evaluated them using RMSE, MAE and RMSE/MAE. As a result, RMSE and MAE were the high values in November to April and the low values in May to October. RMSE/MAE values are higher than values follow normal distribution. Accuracy of this model which predicts monthly rainfall was higher in the rainy season than in the dry season but in the rainy season, many outliers were predicted.

 Column experiments with mixed various particle size distributions were conducted in order to elucidate the influence of physical and fluid properties such as particle size distribution, porosity and Reynolds and Peclet numbers on the solute spreading in Darcy and non-Darcy transition regions. Temporal moments based on breakthrough curves were utilized to estimate the “apparent dispersivity”. It was confirmed that the particle size distribution affected seepage flow regimes and subsequent apparent dispersivity estimates. In addition, linear multiple regression analysis using AIC was performed on physical and fluid properties that had a correlation with the apparent dispersivity. Furthermore, a new estimation function involving the uniformity coefficient and Reynolds and Peclet numbers associted with Darcy and non-Darcy flow regimes was proposed, resulting in a good agreement between experimental and estimated results.

 In a river running on permeable bed, stream water is exchanged with riparian aquifer by infiltration or exfiltration. When groundwater level is lower than stream water level, seepage flow occurs through its riverbed layer to the aquifer. Then if groundwater level is higher than riverbed, the seepage becomes a saturated flow, and a practical formula to evaluate the seepage flux has been established conventionally. On the other hand, in case that groundwater level is lower than riverbed, the seepage becomes an unsaturated flow, and it is known that the conventional formula of seepage flux gives underestimate values. For this situation, though authors already proposed an alternative formula in the past, its application range of the formula was limited to the case of deep groundwater level. In this research, a more versatile formula to be applicable to the case that groundwater level is closer to riverbed is theoretically derived, and its effectiveness of the new formula is verified based on numerical simulations of unsaturated flow.

 River water is lost on upper-area of alluvial fan and gained on lower-area. That is a natural flood control function. I estimated the Water Storage Volume in an alluvial fan by using a numerical simulation model that can simaltaniously solve surface water and groundwater flow. The groundwater transmissivity is given as 1.0 m²/s and the water movement of entire alluvial fan was computated. As a result, the peak discharge was reasonably calculated as 4,266 m³/s against observed discharge 4,413 m³/s. Its maximal Water Storage Volume was 66 million m³ and sufficiently large compared with flood control volume of dams. Then the alluvial fan contributes small to the peak-cutting of flood discharge, but the storage effect can continue during flood with multiple peaks. That feature of flood control is unique not to dams but to alluvial fans.

 In groundwater simulations, an inverse approach for the least square errors between measured and calculated hydraulic heads is not necessarily valid when the head data include uncertain measurement errors as in most cases. The purpose of this study is to compare agreements of true hydraulic heads at all nodes with estimated heads after the inversion with uncertain head data. This study applied initial parameter settings, matrix-combinations, traditional least-square optimization, and genetic algorithm for 5000 different parameter problems in each one-, three- and five-layered 2D steady flow simulation. As a result, the initial parameter setting provided the least mean squared errors for all nodes most frequently, indicating inversion procedures were not necessarily appropriate when the calibration data included small but random errors. The linear discriminant analysis showed the appropriate inversion for all nodes could be determined with a high prediction ratio around 0.8 although inversion results in all four methods were required.

 Spring water which appears in the slope during heavy rainfall is widely known as one of the precursors of slope failure. On the other hand, conventional seepage analysis assuming the uniform field cannot express the condition of locally appeared spring water. In addition, during heavy rainfall, the pore air pressure increases with increasing saturation near the ground surface, and as a result, rainfall infiltration is suppressed, but this phenomenon is not considered in conventional seepage analysis. In this study, the mechanism of spring water and the effect of pore air at the time of heavy rain on non-uniform slope were examined by three-dimensional numerical simulation using a spatial distribution model of hydraulic conductivity.

 A distributed runoff model is composed of physically based element models and it is expected as to be applied for forecasting unprecedented flood disasters. However, vertical unsaturated flow in mountainous area has not sufficiently modeled in many distributed models. In this study, we observed water content and pore water pressure at Kiryu experimental watershed. We measured the water content in soil and bedrock layers. Based on the analytical solution of Richarads’ equation with diffusive wave equation, we represent the observed pore water pressure changes.We revealed that water flux derived by the analytical solution only with a diffusion coefficient and we comfirm appropriate representations against the observations.

 In recent years, extreme weather has been exacerbating on a global scale. Therefore, heavy rainfall has been frequent in many places. This is no exception in the metropolis, and people will suffer from river floods and rainfall inundation if rains are unprecedented. In this study, Yokohama city and Kawasaki city, which contain the Tsurumigawa river basin, were taken as the objects, and the actual rainfall data observed on August 22, 2016, by the X-band MP radar of the ministry of land were used for reproduction calculation. Although precision verification has been carried out in the 23 wards of Tokyo before, the comparison between the result of this reproduction calculation and the water depth and river depth at that time can confirm that S-uiPS is also applicable in other cities. Furthermore, the calculation includes a part of gutters in some areas of Kawasaki city in this study. In the future, S-uiPS can also be applied to moderately urbanized areas, and its usage method is also described in the paper.

 Recently, large scale water disasters which large amount of water flows into building have occurred. Therefore, it is considered the importance to develop an inundation analysis model considering inflow and outflow into building for examination of flood control. In this study, based on the inundation analysis on Cartesian coordinate system, analysis model was developed with the hydraulic model of the building considering inflow and outflow. The characteristics of the analysis model were shown and the validity was confirmed from the viewpoint of mass conservation. Moreover, as the application of the analysis model, the effect of building buildup on inundation, the amount of economic damage of houses, the effect of the destruction of door and windows were examined.

 In this research, the simulation of flood in the Kagetsu River in July 2017 was conducted using 2D flow model in order to evaluate the effect of bridge beam on flood capacity, water level and flood scale. The flood flow in urban area was also simulated.

 As a results, bridge piers increased inundation velocity and inundation area. When water level exceeds the height of bridge beam, flow discharge was limited by the area of bridge opening, which caused flow capacity decreasing.

 In recent years, inundation damage caused by heavy rain has occurred frequently. Numerous researchers have developed numerical analysis models for inundation due to heavy rain. But, observation data used for verification of analysis models are insufficient. In this study, water level gauges were installed in the sewer system of the Jizo river basin in Kasugai city, Aichi prefecture, Japan, and water levels of the sewer were observed. The effective rainfall was defined for the examination of the observation results. It is shown that the peak value of the effective rainfall is highly correlated with the peak value of water level. In addition, in order to examine countermeasures of inundation due to heavy rain, numerical analysis of urban inundation considering drainage facility and water storage tank was developed. The accuracy of the inundation analysis model was verified from the comparison between analysis results and observed results. Moreover, the decreasing of the maximum inundation depth and the water volume on the ground were confirmed by numerical simulation with drainage facility and water storage tank.

 It is necessary to input rainfall data of spatially high resolution into a flood model for assessing urban flood risk precisely. However, the extent of the required spatial resolution of rainfall data was not fully investigated.

 The aim of this study was to evaluate the impact of rainfall distribution caused by wind-driven rain (WDR) in an urban area on inundation. At first, wind distribution was given by RANS model. Secondly, the movement of wind-driven rain was simulated and the distribution of rainfall was obtained. Finally, inundation simulation was conducted.

 The simulation results showed that the distribution of rainfall intensity at drainage areas changed dramatically with the increasing wind speed. Flow rate on drainage areas of which size is less than 200m × 200m could increase more than 20% by the wind. Therefore, it is suggested that the required spatial resolution of rainfall data in an urban area was less than 200m × 200m.

 We examined applicabilities of a nested approach to flood simulations with non-rectangular domain. In the case of rectangular domain, setting of nested domain has difficulties because of locations of land elevations and rivers. To solve the problem, nested calculation approach with non-rectangular domain which was applied for unstructured flood model was adopted. Gaibandha District which is located in north part of Bangladesh was selected for our study area, and Udakhali union was selected for the nested calculation area. We used observed water levels, and 121 days calculation was conducted. As a result, we confirmed that there is no serious influences to simulation results due to the difference of nested domains. Also, we were able to save the calculation load with the setting of the un-rectangular domain.

 In recent years, fluvial flood damage has been mitigated by strengthening embankments, however, inundation by heavy rainfall is increasing in mega and provincial cities. Fukuchiyama is one of provincial cities and prone to fluvial flood. This city was frequently attacked by Yura river floods and damaged severely in years past. And then, pluvial floods in 2014 and 2018 brought hole city inundations. The drainage system includes small tributaries and those river water is pumped out. This system causes city inundation if pump stations can not work because water level of main river is high. In this paper, this kind of pluvial flood and safe evacuation are discussed. From the results, inundation characteristics and safety of evacuation center are clarified by using 2D shallow flow model.

 The accurate flood forecasting technique is required to provide information of flood evacuations for residents in an inundation area. The run-off analysis in the flood forecast technique is usually verified by rating curves of rivers in general. But the rating curves can not accurately estimate discharge hydrograph in rivers where effects of flood storage and bed variations are large on flood flows. And the runoff from the small basins is not sufficiently verified. In the present study, the flood flow and bed variation analysis using observed water surface profiles provides accurate discharge hydrographs of the Hii River and the tributaries of 2013 flood. And parameters of the tank model are identified by using those discharge hydrographs. The flood forecasting technique using the verified tank model accurately predicts the time series of water surface profiles and discharge hydrographs of the 2011 floods in the downstream of the river.

 Estimations of flood flow discharge and model parameter optimization such as roughness coefficient were performed by flood inundation analysis using water levels as boundary conditions for the Kagetsu river during North Kyushu heavy rain in 2017. The optimization of model parameters of rainfall-runoff and flood inundation flow analysis was also performed based on the estimated flow discharge. It showed that the method proposed in this study enables optimization of model parameters of rainfall-runoff and flood inundation flow analysis even when flow discharge was unknown, and also was confirmed to improve the reproduction accuracy of flow and water level hydrograph while maintaining the same degree of accuracy in the maximum flooded area.

 Numerical simulations of overland, flood-inundation, and groundwater flow in the Akatani river basin during Kyushu-hokubu heavy rain in 2017 were performed by a dynamic model which is based on two-dimensional shallow water equations and the free-surface groundwater equations with water exchanges due to infiltration processes. The model was also verified against observed flood marker and the maximum inundation area which showed that the proposed model has reasonable accuracy. The effects of infiltration and groundwater flows were also examined based on the numerical results.

 We developed numerical simulation models for water flows in the agricultural drainage systems in Kamedago district in Niigata City, which is located in one of the low-lying agricultural areas in Japan. Then we examined the effect of preliminary operation of drainage pumping stations and lowering the target water level of drainage pumps on suppressing the peak of interior water level and mitigating inundation damage during extremely heavy rain. As a result, it was revealed that the suppression effect of the peak of interior water level is effective especially in the early stage of the rainfall event, and that interior water level can be continuously lowered by lowering the target water level of drainage pumps. In addition, as a result of the examination by using the probable rainfall hyetograph, the inundation damage was found to increase when the peak of the rainfall is located later during the rainfall event. It is also implied that the effect of lowering the target water level of drainage pumps on mitigating inundation damage is particularly effective for the paddy fields, and larger as the peak of the rainfall is located former during the rainfall event.

 Damage by a typhoon and a local torrential rain occurs frequently at a various part of Japan. The appropriate measures to each water disaster that may occur in the future must be requested. In this research, assuming that a large-scale torrential rain and a storm surge occur at the same time, we made a flood prediction calculation by using flooding forecasting method S-uiPS. The flow which goes upstream a municipal river from coastal region is caused by a storm surge. As a result, even though the spot away from the river-mouth, there is the risk of the occurrence of the flow backward from river to sewage and the overflow from a river . This research is aimed to clarify such risks more clearly.

 Recent global warming and resultant floods have brought the frequent inundations. With this phenomenon, the real time inundation forecasting on a national scale has become desirable, containing not only large size rivers but medium and small size rivers with only a few available observational data. Until now, flood forecasting has primarily been focused on water level of streamflow. But now, the practical application of grid cell-based models for integrated analysis of rainfall-runoff and inundation is in progress. In this study, the Rainfall-Runoff-Inundation (RRI) model is applied as a dominant model capable for calculation for inundation forecasting on a national scale. We succeeded in construction of wide area models for respective regions based on the Japan Flow Direction Map, realization of high speed by improvement of vector-type super computer program corresponding to parallelization, realization of inundation forecasting in real time and construction of a Web displaysystem for all results. In this way, the national scale platform for inundation forecasting system has successfully been proved, while individual examinations are also necessary for improvement of accuracy on respective rivers.

 The effect of inundation information sharing among evacuees on flood evacuation is quantitavely estimated. The micro flood evacuation model is modified so that it can simulate the case where evacuees can share the water depth and location of inundation which they come across with all other evacuees, and based on the shared information evacuees select new routes to shelters avoinding inundated places. The model is applied to the Seri river douwnstream area in Shiga prefcture and how much the information sharing can reduce the failures are estimated.

 It is important to start evacuation early to avoid encountering river floods. However, in reality there are few people who start evacuation according to evacuation advisory. In addition, it is impossible to predict the failure point and time of river embankment during heavy rain. Therefore, it is also important to predetermine evacuation routes that are unlikely to encounter floods. In this study, we will establish a route selection method to minimize casualties in evacuation from river flood. And we compare how many shelters are equivalent to the effect of reducing the number of casualties. The calculation results show that it is possible to reduce the number of casualties compared to the shortest route to the refuge by route selection to avoid the flood taking into account the flood distribution in the area. The optimal route which minimizes the number of casualties is obtained from the distribution map which acquired the minimum value of inundation start time of each flood scenario, and it turned out that the casualties can be reduced compared with the shortest route selection in all antesipated floods. It was found that the reduction effect of the casualties by the optimal route is equivalent to the installation of 25 additional refuges.

 A two-dimensional flood model was applied to the middle reach of the Arakawa River (AR), to clarify the change in the potential flooding risk area, under three breaching conditions including the breaching by infiltration (BI). The water level tends to be high at the confluence of the main river and the branches and the change point of the riverbed slope. The high risk locations were at the upstream of the AR, at the junction of AR and Oppe river, and at that of AR and Iruma river. Under the influence of breaching of the upper reaches of AR, breaching timing of the second embankment became early in Yoshimi Town. Kawajima Town (KT) has a topography that is easy to be flooded.As Sakado City and Kawagoe City, outside the KT, are also likely to be flooded, it is effective for the people in KT to evacuate to the left-bank side of AR. In the southeastern area of KT, the inundation depth is high and there is a danger of house collapse. Evacuation to the left bank of AR is recommended from that point of view.

 The paper deals with analysis of a large-scale evacuation drill by Ashiya high school. Ashiya high school carried out the evacuation drill against Tsunami with more than 1000 high school students on December 21 2016. First, authors took the GPS log of the evacuation drill and clarified the basic evacuation behavior of the students. Then, the GPS log was compared with the newly constructed multi-agent evacuation simulation model. To make the simulation model more accurate, the velocity equation of the evacuee was newly constructed using a fuzzy inference theory. As a result, the multi-agent simulation model was able to reproduce the GPS log more accurately.

 For flood risk assessments, a risk rank was proposed. Risk rank is defined by using inundation depth and inundation flow velocity. In this study, we evaluated the risk rank using the results of flood simulation for the Kakehashi River basin. Through the calculation process of risk rank, it is possible to estimate the time limit of evacuation, or the time when people cannot evacuate on foot. Then, we calculated the possible evacuation time as the difference between the time of evacuation limit and of the announcement of evacuation recommendation. The risk rank is defined by its maximum value. On the other hand, the possible evacuation time, or the degree of evacuation difficulty, is affected by difference in temporal pattern of flood. In this study, evacuation was more difficult in the case of lower risk rank. These results indicate that the larger flood event does not always cause the difficult condition for evacuation, but the smaller flood could make evacuation more difficult. Evacuation activity is quite important to reduce human loss in the case of large-scale flood which brings inundation by breaked levees. Therefore, it is equally important to estimate both the risk rank based on the maximum inundation depth and the possible evacuation time by considering various patterns of flood. Possible evacuation time can be estimated for individual area in the target region. Thus, it is clear and useful information to decide when each person should start evacuation activity to escape safely.

 In recent years, large-scale floods that caused many fatalities occur frequently in Japan. It is necessary to grasp accurate flood risk under future climate for considering adaptation plan. The LIFESim model which estimate the number of fatalities by water depth in flooded area is widely used in Japan. However, the “Floris model” which estimate by not only water depth but fluid force, water velocity and water rise rate has been used for taking measures on climate change adaptation in the Netherlands.

 In this study, we estimated the number of fatalities by LIFESim model and “Floris model” for the heavy floods under historical and +4K future climate based on the large ensemble dataset (d4PDF) and clarified the differences of the numbers of fatalities by each models.

 In recent years, pluvial floods by heavy rainfall are increasing and sometimes cause underground inundation in urban areas such as Osaka, Tokyo and so on. Subway stations are also prone to this inundation, and the intruded flood water into subway spreads through subway tunnels if counter measures to the inundation would not work well. In this paper, evacuation lead time for subway users is investigates by using numerical model including drainage system, ground and underground spaces. Number of flood sufferers in subway is estimated numerically and vulnerability of subway station is discussed. From the results, evacuation lead time and vulnerability to inundation of each station are pointed out. These results are important for government officials to make an evacuation plan from subway stations.

 In order to understand the important factors that affect evacuation behaviors of local residents in the heavy rain event of July 2018, a principal component analysis on evacuation awareness and disaster prevention information access was conducted using a dataset of questionnaire survey for local residents near Nagaragawa River in Gifu Prefecture. The first principal component was the negativeness index of evacuation during the heavy rain event, indicating that many residents were reluctant to evacuate for groundless reasons. The second principal component was the utilization index of disaster prevention information during the heavy rain event, suggesting that during the decision making, non-evacuated residents could be divided into those who decided not to evacuate based on the information and those who did not. Evaluating the averaged properties for each of the categories divided into four quadrants in a biplot, it is shown that each type of the non-evacuated residents had inherent recognitions to evacuation and information.

 Large flood disaster occurred in Mabi-town, Kurashiki, Okayama Prefecture, due to Western Japan heavy rain in 2018.43 of 51 deaths were found on the first floor. In this study, to understand the potential risk of inundation in a house, we conducted a full-scale inundation experiments. Here, a full-sized room and a door were made in the water tank, and the door opening the furniture overturning/scattering, and the evacuation experiments were conducted. As a result, it was confirmed that the door could not be opened even if the water-level difference on both sides of the door was only 15-20 cm. The results of the furniture overturning/scattering experiments show that when the water level is high, the closet may fall down and there is a risk of injury and the evacuation passage may be blocked.

 Using the questionnaire results for the residents in Mabi town, Kurashiki city, Okayama Prefecture, which suffered massive damage in July, 2018, we compared the awareness of the evacuation behavior and the hazard map. As a result, the degree of recognition of the hazard map is over 20% for both types of questionnaires and is never high. From the questionnaire results of Okayama Broadcasting, the higher the awareness of hazard map is, the higher the level of understanding for disaster information such as heavy rain special warning and evacuation instruction is. Furthermore the higher the awareness of hazard map is, evacuation behavior is earlier.

 The overflow breaches of the river levees in large-scale experiments and a historical flood event have been investigated using a process-based model with considering grain-size distribution of levee materials, for evaluating the validity of the numerical model on the hydro-morphodynamic process of river flood, levee failure and inundation. The comparisons between the numerical and experimental results show that the model can reproduce detail processes of levee overtopping, breach deepening and widening, and flood-plain erosion and accretion measured in three different experimental conditions. The multiple grain-size model has an effective advantage in simulation of lateral breaches of the levees exposed to high-velocity flood flows. The application of a real levee-breach event on the Ikarashi River (Japan) in July 2004 shows that the model with detailed hydro-morphological conditions is appropriate for the estimation of final breach width, morphological changes in flood plain, and inundation volume.

 In recent years, the damages of river levee have been occurred due to heavy rainfall, and it is possible that the number of damages will increase with the increasing of the heavy rainfall frequency. Therefore, it is important to find weak points of a river levee in order to efficiently establish the countermeasures due to heavy rainfall. The purpose of this research is to grasp the ground composition of river levee which are considered to be danger due to the change of rainfall frequency and to determine the dangerous point of the embankment based on the analysis of future climate data. In order to solve this problem, a numerical analysis was performed to determine the dangerous point in the embankment due to future climate change. Therefore, to show the validity of the analytical model to be used, a verification test was carried out under seepage condition by using numerical analysis. In order to grasp weak points of river levee, parametric analysis focused on hydrograph and ground information was carried out. At three locations in Chikugo River, the seepage flow analysis was carried out using hydrographs based on future climate data.

 In the Tokachi River large flood in 2016, there was no significant damage to the levee in the downstream area where the gently sloping levee has embanked. It means that the gently sloping levee has high stability not only for the soft ground measures but also for the seepage flow during a flood event. In this study, the downstream of the Tokachi River is targeted, the history of earthquake disasters and levee maintenance are examined, and the soil quality and structure of levee are grapsed. Then, using the water level hydrograph, the cross section shape of the levee and the average pearmeability coefficient, the longitudinal distribution of the levee vulnerability index is calculated. We clarify the reason why the gently sloping levee has not caused the seepage failure due to the 2016 large flood based on levee vulnerability index t^*. It is concluded that the wide and gentle slope levee is very effective for seepage failure reduction. Also, the importance of embanking the levee with a structure whose t^* is less than 0.01 is discussed.

 In August 2016, the tributaries of Tokoro River had a levee breach caused by the backwater of Tokoro River. This study aims to clarify the following point, the effect of backwater on levee breach processes by using a numerical model to simulate levee breaches. The results of the simulation are as follows : river width and normal flow affect mechanism of widening of the levee breach opening.

 This study aimed to clarify the phenomenon of ice jams generated around bridge piers. We conducted an ice jam experiment. The section of this experiment waterway is compound cross-section. Polypropylene and actual ice were used for the ice model in this experiment. The study clarified the following. (1) By ice jams, the speed of the ice model decreases and the number increases. (2) The ice model was detoured at bridge piers and flowed on the high-water channel. And it was flowing into the low-water channel in the downstream. (3) After ice jam occurrence, Flooding begins from the upstream site. Later, Flooding flowed to the downstream site. (4) The flooding speed is fast when discharge is large. As time passes, the flooding area becomes the constant value. (5) In the ice jams site, the upstream water level went up, the downstream water level descended. The downstream water level has been increased suddenly after ice jams breaking.

 22 of 94 bridges were damaged in the Misasa River due to the 2018 Western heavy rainfall. The purpose of this study is to propose an index suitable for bridge damage assessment under heavy rainfall based on the longitudinal distribution data of bridge damage in the Misasa River conducted by the authors. As a result, it became statistically clear that the bridge damage characteristics can be explained by using the dimensionless cross-sectional area (= cross-sectional area / catchment area) and bridge index (= bridge area / cross-sectional area) . The evaluation procedure of bridge damage risk using these indicators was also examined on a trial basis.

 In 2016 August, Otofuke River, located in Tokachi Prefecture of Hokkaido, faced active bank erosion leading into a significant migration in the path of the channel due to the four majour flooding events. In this study, following our previous study, the authors focused on the effects of the predicted future flooding scale (it becomes 1.5 ~ 1.7 times larger than present flooding discharge) on channel morphodynamics in the steep-graident rivers. The results showed that as the flow rate increased, the channel width also increased. Moreover, when the flow rate exceeded 1.5 times larger than that of 2016 flood, meandering channel changed their channel formation to multi-thread, braided channel, leading into channel bifurcation and new flow path in the protected inland area. The important things in the steep-gradient river would be to make banks (embankments) resistance to erosion as a river channel planning.

 Woody debris increases the destructive force during flood events, especially their accumulations at river bridges lead to the increase of flooding areas. Furthermore, recent serious flood events have highlighted the need to retain woody debris directly in larger rivers, and to construct the effective retention facilities which have enough capacity for accumulated woody debris. This study aims to evaluate the effect of retention facility of woody debris in river bends by using the simulation model composed of depthaveraged flow analysis and extended distinct element method for woody debris behavior. Results of calculations show the good performance of the retention facilities around river bends in the Seki River and we can recognize the proposed model is useful for the design of the retention facilities.

 During August 2016, Typhoon 10^th has approached to Hokkaido thus causing record-breaking rainfall. At upper stream area of Sorachi River Basin, Minamifurano was inundated due to the overtopping levee at the upper reach and lower reach. The research aims to predict the possibility of river water level distribution, riverbed fluctuation distribution and flood disaster beforehand. This can be conducted by using the weather forecast model which predicts the typhoon path two days prior and using the data to calculate the rainfall-runoff. Then, using river flow calculation model, the flood depth and flood area uncertainties can be predicted. As a result, two days prior, not all ensemble’s calculation result resembles the real flood disaster event. Majority of the ensemble result were not flooded and also there is an ensemble case which inundated worse than the real disaster. Thus, flood possibility and also flood damages of the event were assessed.

 In recent years, river disasters associated with extreme heavy rainfall have frequently occurred, and driftwood that has flowed out from landslides and bank erosion has been captured in the bridge piers, which is one factor that increases flood damages. In Omoto River, Iwate Prefecture, the development plan of the driftwood capturing facility using a channel curving section in a mountainous area was set up for mitigating of driftwood damages in urban area taking advantage of driftwood damages by the typhoon No. 10 heavy rain in August 2016. At present, a structural specifications of the capturing facility that is able to capture the driftwood efficiently by grasping the behavior of the three-dimensional driftwood by hydraulic flume experiments are set. Therefore, in order to analyze the behavior of three-dimensional driftwood obtained by hydraulic flume experiment, the numerical analysis was carried out by applying a numerical calculation model incorporating a driftwood model using the discrete element method to a calculation model of quasi three-dimensional flow. Results suggest that the driftwood in the depth average entered into the catchment reservoir, and there are two types of driftwood such as the driftwood flowing to the slit and the driftwood deposited in the inner band of capturing facility on the bottom flow velocity.

 A two-dimensional flood model was applied to the middle reach of the Arakawa River basin to clarify the change in flooding risk area before and after the change of Arakawa River course. The results were also compared with the present situation and the breaching location in actual flood in 1910. Before the change of river course, flood inundation was occurred in both river side of Kumagaya Fan, and flood inundation was controlled in the Ayasegawa River basin using Bizen Embankment. After the change of the river course, flood inundation to the left side of the Arakawa River was reduced. On contrary, around a river confluence with branches (Wadayoshino River, Ichino River, Iruma River) with the Arakawa River, overtopping from river embankment was increased by backwater phenomenon. It changed Yoshimi and Kawajima area to flood prone area. Especially in Yoshimi, second embankments were constructed inside the Yoshimi Polder. Flooding risk was reduced in the Iruma river basin by the construction of separation levee, however branches of the Arakawa River still has similar flood risk locations after the river course change.

 The paper presents a two dimensional coupled one dimensional numerical model to simulate inundation and sediment deposition over the paddy field considering the effect of channel network. The results show that the roads in the research area have significant influence to flood propagation and sedimentation; Main suspended load deposition happened near the inflow point and along the flow direction, the maximum deposition thickness on the ground surface is over 50 cm. And the sedimentation mode, that is the main deposition area and deposition thickness in the channels is similar to that on ground. The capacity of channel network and crop harvest may be seriously affected by the deposited sediment.

 Situation of landslide disasters caused by the heavy rain event of July 2018 was studied by examining investigation reports and internet information. Occurrence date, time and location of landslide disasters were thereby identified. Landslide Occurrence Possibility of those landslide events was assessed using the Landslide Information Delivery System that utilizes rainfall data obtained by XRAIN, High-resolution Precipitation Nowcasts (NC) and Radar-raingauge Analyzed Precipitation. Comparison study revealed that combining XRAIN and NC contributes to the improvement of accuracy in assessing landslide Occurrence Possibility.

 For effective emergency responses to flood disasters, estimating inundation depth distributions is crucial on the real-time basis. In the previous studies, the authors proposed a method to achieve this objective by assimilating preliminary conducted multiple inundation simulations and local flood related information. This study further improves the applicability of the proposed method with respect to utilzing existing inundation simulations conducted for hazard mapping in all over Japan. Based on the application to the recent flood disaster occurred in the lower part of the Odagawa River basin in Okayama Prefecture, this study showed the reasonable performance for simulating inundation depth distributions and estimating flood volumes even with limited local information obtained at four to eight points with the relative errors of 7% to 3% in the flood volume estimations.

 In July 2018, the heavy rainfall in western Japan caused massive damage to Mabi-cho, Kurashiki, Okayama Prefecture. A large area was flooded within a short time after the onset of dredging. While there is concern about the intensification of heavy rainfall disasters, it is required to quickly grasp the inundation situation. This is expected to use earth observation satellites equipped with Synthetic Aperture Radar (SAR), and the implementation of the emergency observation immediately after the disaster and the development of base-map are in progress. In this research, we propose a method to estimate the inundation depth easily by extracting the flood zone automatically using SAR satellite data of emergency observation and base-map data of the same season in the near past.

 Based on the observation data such as the rainfall and flood in the Takahashi River basin, and furthermore the field survey data on the inundation depth, sediment deposit and others in Mabi-cho flooded area caused by the West Japan torrential rainfall in July, 2018, the actual state and characteristics of the sediment deposit and the density of inundation water in Mabi-cho flooded area are examined. It is illustrated that (1) the thickness of the sediment deposit is less than 1% of the maximum inundation depth at each survey point, which has a tendency to be thinner than that due to tsunamis and (2) as leaving the broken point of river embankment, the thickness of the sediment deposit becomes gradually thinner. The cross-sectional average concentration of sediment in the inundation water and the density of the inundation water are estimated by introducing bold asummption on the three-dimensional shape of the flooded area and bold sediment deposit model, and then using the data on the sediment deposit obtained by the field survey. It is judged that the estimated density is smaller than that of the salt water.