In Japan, the flood management plans for the main sections of first-class rivers are formulated for rainfalls with return period from 100 to 200 years. On the other hand, for other than the main sections, the flood management plans are designed for smaller scale rainfalls. In the case of large-scale heavy rainfall in a wide area, the risk of flooding is higher in tributaries. In the flood inundation area maps produced by the Ministry of Land, Infrastructure, Transport and Tourism (MLIT) and local governments, inundation of tributaries is not considered, and the risk of large-scale floods may not be properly expressed. In this study, an inundation simulation model was improved to express flooding from both a main river and a tributary. The model was applied for the Kakehashi River Basin in Ishikawa Prefecture, Japan, and its tributary Hatcho River. The influence of the levee breach in Hatcho River was small for the maximum inundation depths. On the other hand, the inundation duration time was prolonged more than 9 hours in wide area when levee breaches occur both in Kakehashi River and Hatcho River.

 In the Kakehashi River Basin in Ishikawa Prefecture, inundation simulations were conducted by using rainfall data under the global warming condition. After identifying the areas requiring evacuation based on the inundation simulations, the number of evacuees to each evacuation center was calculated in consideration of the difference in population distribution by time and period. The number of evacuees increases more than 2000 in areas where the population increases during the day. During summer vacation, total number of evacuees can increase. Not only the number of residents, but also the number of visitors during daytime has to be considered to make evacuation plan. In addition, number of evacuees were calculated under urban structural changes assuming relocation from areas with higher flood risk to lower risk areas in the condition of future population decline. Under the declining population, it was possible that the number of people requiring evacuation could decrease 1000 in significantly. To take advantage of surplus area to reduce the risk, it is valuable to take measures to relocate people from higher risk to lower risk areas.

 This study aims to obtain basic knowledge on the hydraulic characteristics of flood flows under storm conditions in order to examine the risk of simultaneous occurrence of floods and storms in rivers due to super typhoons. We conducted simple numerical experiments for flood flow in straight channel with uniform cross section under various wind speed and directions. The results indicate that after the onset of wind stress, the flow velocity in the main flow direction changes abruptly, reaches a maximum value, and then transitions to a steady state. The response of flow velocity to wind stress was faster than the response of water level due to the large vertical mixing modeled with vertical kinematic eddy viscosity.

 Superimposed inundation simulation was carried out for the Gobo River, a second-class river in Takamatsu City, Kagawa, Japan, under the condition of 1000 years probable precipitation, taking into account the sewerage system, for a compound water disaster in which rainwater and river water inundations occur at the same time. As a result, even when the sewerage was taken into consideration, the inundation area became the largest during the superimposed inundation. It was also found that by considering the sewerage, the inundation depth decreased, but there were places where the flow velocity increased. In addition, in the case of superimposed inundation, the results showed that a wider area was inundated than simply adding the inundation areas of inland waters and foreland waters.

 A model was developed to accurately assess the risk of waterlogged inland flooding caused by the inability to drain rainfall due to the closure of sluice gates caused by rising river water levels. The analysis was conducted for Setagaya City, Tokyo. It was found that if rainfall increases by a factor of 1.13 to 1.15 due to climate change, the area inundated above floor level increases by a factor of 1.22 to 1.27. Comparing multiple rainfall patterns, the results show that at the end of this century, up to 2.9 m of inundation occurred during a 200-year rainfall event. It was also found that the damage caused by 1/100th and 1/150th of the current rainfall would increase to 1/50th of the current frequency by the end of this century. Furthermore, it was shown that the social impact would be further aggravated by prolonged sluice gate closure times.

 Recent years, heavy rainfall disasters have been more frequently occurring due to climate changes, and the damage reports from complex disasters caused by debris flows increased. In this study, debris flow simulations were conducted at 171 locations along railroads in Yamaguchi Prefecture to evaluate the risk of complex disasters in which debris flow can flow into railroads and rivers based on sediment deposited thickness. As a result, there were many places where sediment flowed into outside the debris flow hazard area in the simulation. In addition, the railroad damage risk was higher in the places where the railroad was located at the foot of the mountains and where debris flows contacted the railroad after long distance move. The results also suggest that the risk of complex disaster can be assessed by a statistical model with topographic features that correlate with the complex disaster risk estimated from the simulations.

 “Robustness, Recovery after damage, Emergency response and Build Back Better” are key issues to achieve disaster resilient society. Torrential rainfall in July 2018 affected the enterprises in wide area of the west Japan, bringing direct loss due to inundation and landslides, and indirect loss due to power outage, water outage, road blockage, etc. This paper conducted an questionnaire survey to the affected enterprises in Hiroshima and Okayama Prefectures to understand the actual situation of recovery of their headquarters, branch offices and factories. Recovery curves of their activities and difference of these curves due to direct or indirect loss were identified.

 Future rainfall projections based on a large ensemble data assuming a future climate have shown that heavy rainfall tends to be spatiotemporally concentrated. In this study, Line-shaped rainbands (LRB) was defined as a linear and stagnant rainfall event based on the horizontal 5 km resolution of radar-raingauge analyzed precipitation products, and the method was applied to d4PDF, a large ensemble climate database that includes both current and future climate. The ensembles with the top 5% of the number of LRB in the past/+2K/+4K experiments show that warm moist air from the southwest is dominant around Hokkaido due to the strong Pacific High in many LRB years. On the other hands, in the year with the bottom 5% of number of LRB in each experiment, Okhotsk high is stronger and warm moist air from southeast is few. This result indicated high universality of the relationship between synoptic conditions during summer and extreme occurrences of LRB.

 Typhoon Hagibis in 2019 induced river floods in various parts of eastern Japan. The precipitation of Typhoon Hagibis was reproduced by a high-resolution atmospheric model experiment, and the model showed high reproducibility. Atmospheric model experiments with the pseudo-warming method, in which the environmental air temperature increases are assumed to be 1.1 and 3.4 degrees, were applied to estimate precipitation, where the future typhoons show almost the same intensities and tracks. The increases in rainfall, which was mostly comparable to the effect of increasing water vapor, were estimated in the future typhoon experiment. Next, the river inundation in Ibaraki Prefecture was reproduced with an RRI model experiment, and its climate change responses were also investigated. It was projected that the peak discharge would increase, the peak water level would rise, the maximum inundation depth would rise, and the inundation area would expand. In particular, there was a significant difference in the impact of flood influences between the cases where the environmental air temperatures increased by 1.1 and 3.4 degrees.

 Sediment discharge is said to increase due to heavy rainfalls associated with climate change. We used rainfalls from d4PDF, a large-ensemble climate simulation database, as input for a numerical model to calculate flow and sediment discharge in the upper reaches of the Pekerebetsu River in Hokkaido, Japan. Our simulation results showed that the increase rate of sediment discharge from present to future climates was much larger than that of rainfall and flow discharge at higher non-exceedance probabilities. We also found that climate change causes a greater increase in sediment discharge from slope failures than from riverbeds.

 This study aimed to analyze the erosion risk based on time-series changes in flow rate and riverbed changes of rapidly flowing rivers under climate change. The Toyohira River, the subject of this study, is a rapidly flowing river that flows through the urban area of Sapporo and is prone to erosion due to high-speed flood flow. In this study, in order to take into account the uncertainty of climate change, runoff calculations were conducted using a large amount of ensemble rainfall data obtained from d4PDF. Next, a riverbed fluctuation calculation was conducted for each event to estimate the time-series changes of riverbed fluctuations and erosion risk. Finally, a cluster analysis of the present and future spatial distribution of rainfall was conducted to determine the spatial distribution of rainfall that is likely to cause significant flood damage. The results show that the magnitude of erosion risk varies temporally and spatially depending on the characteristics of the runoff, and suggest that erosion risk tends to increase when the rainfall is concentrated in the middle part of the Toyohira River.

 This study projects long-term change in water resources over Japan using the 150-year continuous run of the Meteorological Research Institute's climate model (MRI-AGCM3.2S). This study assessed which regions will experience tight water resources by comparing the daily supply and demand of water under the future climate due to the effects of changes in the spatio-temporal distribution of rainfall and continuous rainfall-free conditions caused by climate change. Then, the results revealed that the areas where water resources are expected to be particularly tight are the Tokachi Plain, the Pacific Ocean side from Chiba to Fukushima prefectures, the Matsumoto Basin, the central Kinki region, the Seto Inland Sea coast, and the Kyushu-Okinawa region. There are areas where the use of water resources become tightened although the annual water resource availability increases. That is because the increased amount of water resources can’t be used depending on the water supply-demand balance during the period of increase.

 This paper investigated the potential impacts of climate change on seasonal operation of the Tedorigawa Reservoir in the Tedori River basin, a heavy snowfall basin in Japan, through the simulation of long-term reservoir operation using a single run of 150-year continuous climate experiment projected by MRI-AGCM3.2S of Meteorological Research Institute, Japan, based on RCP8.5 and SST ensemble mean condition. As a result of investigation, it was shown as a possible future trajectory under climate change that non-negligible changes could be seen in seasonal operation of the reservoir in the late 21^st century, due to changes in snow processes as well as those in streamflow after summer season.

 There is growing concern about the health risks posed by global warming and the heat island effect. In order to properly assess health risks, it is important to evaluate them in terms of thermal comfort. This study evaluated the long-term trend of the Universal Thermal Climate Index (UTCI) in Japan in summer from 1980 to 2020, and the relationship between UTCI and the number of heat stroke patients. Regarding the long-term trend of UTCI, increasing trends were detected at 109 out of 140 AMeDAS sites. Thermal comfort had worsen since the rise in air temperatures, MRT, and decreasing wind speeds. Regarding the relationship between UTCI and the number of heat stroke cases, we found that the UTCI threshold at which the number of heat stroke cases drastically increase differs across cities geographically and lower in northern cities than southern cities.

 This study examined the possibility of large-scale inundation occurring in multiple watersheds. The Kanto region was used as a study case, and d4PDF (5 km, SI-CAT), a large-scale ensemble climate prediction data set, was used. Events that exceeded the designed rainfall were defined as rainfall events that could cause inundation, and were compared with other basins during periods when up to three days (two days) of annual rainfall occurred in the reference basin. As a result, in the future climate, rainfall events that generate the maximum annual climate will be prominent in the basins that are close to each other, but they also tend to increase in the entire Kanto region, increasing the possibility of simultaneous flooding.

 The objective of this study is to estimate the risk of slope disasters throughout the year in the heavy snowfall region of Hokkaido, taking into account rainfall and snowmelt, based on the uncertainty of climate change. The Nakayama and Nissho Passes, which are the subjects of this study, are strategic transportation points in Hokkaido, and a road closure due to slope disaster will have a great impact on logistics. In this study, d4PDF ensemble data of rainfall and temperature were used to estimate the slope disaster risk in the Nakayama and Nissho Passes. In addition to the topography, geology, and ground conditions of the target areas, hydrological conditions such as rainfall and snowmelt were taken into consideration, and appropriate corrections were made to the d4PDF to evaluate the risk level. As a result, it was shown that the frequency of slope disasters increased significantly due to changes in the hydrological environment because of the effects climate change, and that some areas show a significant increase due to snowmelt as well as much rain, while other areas show a significant increase due to heavy rain during the summer depending on regional characteristics.

 We propose a new approach to physically quantify the change of snake line patterns with the application of Clausius-Clapeyron scaling law. The key of the proposed approach is to establish the physically based linkage among air temperature change on ground surface, precipitation intensity, and corresponding snake line pattern of certain percentiles of rainfall events. We used 99^th and 50^th percentiles for representing extreme and general precipitation conditions. With the long-term observation at Kobe meteorological observatory, we successfully verified the applicability of the proposed approach. The climate projections of 5km NHRCM were then analyzed to examine the future change of Clausius-Clapeyron scaling in the Japanese archipelago. It is also revealed the change of snake line patterns under climate change influences.

 The changes in future extreme precipitation that will occur due to anthropogenic climate change remain an important subject. Extreme precipitation is proportional to the amount of atmospheric precipitation available, possibly following the Clausius-Clapeyron (CC) relationship (7%/°C). In this study, we attempted to analyze the influence of the application of CC scaling on the frequency change rate of extreme precipitation events with different time scales in Japan using d4PDF past and future climates experiments for +1.5K, +2K, and +4K rise in the global mean surface air temperatures. The results showed that the frequency change rate of the past period was within the theoretical range of the CC scaling, regardless of the precipitation duration. The frequency change rate of future periods tend to fall below the theoretical range of CC scaling as temperature increases. However, result for the most extreme event with 1-hour precipitation in the 4 °C temperature increase experiment showed that the increase in frequency change rate was within the theoretical CC scaling range in the northwestern Kyushu, western Chugoku, Hokuriku, Kanto, Tohoku, and Hokkaido regions.

 We investigated the influence of two different treatments of water vapor in pseudo-warming experiments for a heavy rain event in August, 2021; one keeps relative humidity in the present and calculating future humidity field by temperature increase; the other first calculates the difference of specific humidity between thre present and the future and reflects to the difference of relative humidity. Two different treatments yielded the very different precipitation pattern each other, and weakened the rainfall intencity compared with the present simulation even though equivalent potential temperature increased drastically in future simulations. Atmospheric stability became stable in future cases, which led to suppression of rain.

 Reanalysis data has been used instead of observational data in the case of bias correction method is applied for outputs from GCM in an observational data scarce area, in which observational period is less than 20 years. However, reanalysis data does not match the real weather characteristics in a local scale since it is targeted at a global scale. In previous study, a statistical correction method for precipitation of reanalysis data has been developed. However, the method has not been applied for hydrometeorological data except for precipitation. Therefore, this study also developed and applied the method for the hydrometeorological data (temperature, humidity, pressure, shortwave radiation and wind speed). As a result, it is revealed that the distribution of corrected data by the statistical correction method is almost same relative to observational data. Furthermore, in order to increase the correction accuracy, it was found that we should apply probability distribution function (PDF) which is fitted the distribution of observational data according to the situation, since optimal PDF in the statistical correction method is different by the target station/area and/or hydrometeorological data.

 In previous studies, climate data is considered the uncertainies by multi-ensemble data in climate change impact assessment. On the other hand, impact assessment model such as runoff model has not been taken into account the uncertainties like the range of predicted results by the model using the range of parameters that change with the predicted floods. Therefore, this study aims to evaluate the importance of the uncertainties of the runoff model for the future predictions by using d4PDF as input data to the model which is used the range of parameters ("the possible runoff model"). As the result, it is suggested that the changing trend that there is possible of missing such as the frequency of flood and draught in the case of using only "the best parameters model" which is optimized in calibration period could be able to predict by using "the possible runoff model".

 Understanding the number of juvenile ayu running upstream is crucial for ayu stock assessment. However, counting is currently done by visual inspection of underwater camera images, which is a time-consuming task. In this study, we used images from underwater cameras installed in a fishway at Kamogoshi weir on the Yoshii River in Okayama Prefecture to train a model for detecting juvenile ayu running upstream using YOLOv5 (i.e., a deep learning algorithm for object detection). The model's ayu detection results showed that accuracy improved as the amount of ayu data increased. When labeling an object that is easily misidentified as ayu (i.e., Clithon retropictum), accuracy decreased using small amount of data. Using our model, the number of images requiring visual counting was reduced to less than 1 % of the conventional visual confirmation. Furthermore, using our model in a two-month ayu stock survey is expected to reduce the required time by about 90 hours when compared to the conventional method.

 In recent years, the dumping of garbage in rivers has become a common occurrence and has gradually started to affect the normal flow of river channels, which has added lots of work to the river patrol staff. Facing these problems, river authorities urgently need a reasonable and better cost-performance method, that can be adopted on a large scale to support the staff in investigating the garbage within the rivers. Although object detection using Artificial Intelligence has its advantages, it has not been widely applied in the riverine environment using drone. This study attempts to detect garbage in the Asahi River, Japan using two object detection models. By using a large amount of PET images collected from the Internet as training dataset and experimenting with a variety of model-related parameters (i.e., Batch size, Epochs), this study achieved high-accuracy results in recognition of the garbage in the study site. Conclusively, the additional dataset of PET for training, with the similar GSD as test dataset, can improve the Recall value. Nevertheless, without combining with Original dataset collected from the study site, it is difficult to detect the PET using additional dataset only. Thus, combination of Original and additional dataset is a relatively better method to improve the Recall value of detecting PET.

 In general, a deep-leaning neural network (DNN) poorly predicts without a large amount of data although DNN has been testedly implemented as a quick predictable model into pumping and drainage systems in lowland for management efficiency. Using support vector regression (SVR), we developed a water-level prediction model that can perform good predictions even with a small amount of data. We employed DNN to compare with SVR for an accuracy performance based on data size. The input data are the rainfall and water level, collected in two different lowland areas. The datasets are separated into the short-term-period data and the long-term-period data based on the number of seasonal variations. For the test of the short-term period, SVR predictions upto 6 h lead time were better than those of DNN by 6 to 28% improvement. For the long-term period test, both models had similar performannces in prediction accuracy. In addition, the accuracy of SVR was poor when quick changes of water level were simulated.

 Input variable selection is one of the most challenging tasks for modelers when building a hydrological model using artificial neural networks (ANNs). The conventional method of input variable selection for ANN considers the linear correlation of each variable with the prediction target variable. However, this conventional approach can potentially limit the ability of ANN models. This study surveys the sensitivity of input variable selection methods in ANN performance to obtain an idea to save our time and concerns related to input variable selection. We prepared three ANN models with different input variable selection methods and two regression models as well. Comparing the results from these five models, which are for hourly-based water stage prediction at the Hirakata station, indicates that ANN provides satisfactory prediction accuracy without a careful input variable selection process. And, there is a possibility that ANN performs poorly if the variable selection process eliminates the necessary data.

 Accurate prediction of dam inflow during snowmelt is extremely important for disaster prevention and water use. Currently, AI is being utilized in the hydrological field, and research is being conducted on predicting river water levels and dam inflows. In this study, we used deep learning to predict dam inflows during the snowmelt season on an hourly basis. The results showed that reproducibility was high up to 24 hours ahead, but decreased after 24 hours. Therefore, it is considered that the practical limit of the forecast is 24 hours ahead. In addition, the effect of the number of intermediate layers was more significant than the effect of the normalization and standardization processes.

 Prediction of dam inflow for small and medium-sized rainfall events is strongly affected by the noise of dam inflow, caused by the fluctuations of the reservoir’s water level. In addition, recurrent neural net-work (RNN), which is one of the prevailing deep learning models, have difficulty in adapting the condition of the basin preceding the target rainfall events. From these viewpoints, firstly, we applied Wavelet transforms to reduce the noise of dam inflow during the flood period. Secondly, we proposed encoder attached RNN to adapt the prior condition of the basin. Finally we compared it with multi-layer perceptron (MLP) to verify the accuracy of the predictions up to 24 hours.

 At a multipurpose dam, it is necessary to forecast inflow to control increasingly severe and frequent floods. For more effective dam operation, we developed Maruyama Dam inflow forecasting system using deep learning. To improve forecast accuracy and to decide the structure of the forecasting system, we identified input data that are highly correlated with dam inflows and optimized hyperparameters with a large impact on forecast accuracy.

 The deep learning model has low forecast accuracy for unusual and inexperienced floods because of little training data. Therefore, the system has two applications to improve this problem. One is the forecasting system using the storage function model during the severe floods over the limit of the deep learning model forecasting. The other is to train the unusual and inexperienced floods to improve the range of forecast by the deep learning model.

 DX in the infrastructure section has been gotten attention by recent understaffing problem in the field of civil engineering and the changes of social situation. AI has been widely used in facilities inspection and flood disaster prevention. Although the utilization case of AI increases in the infrastructure section, the method which uses it in the flood and dam inflow prediction has not still been established. Therefore, the purpose of this study is to develop the dam inflow prediction model by using Deep Neural Network (DNN) for the Shiokawa Dam where was constructed in the Shiokawa River of the Fuji River system in Yamanashi prefecture. Furthermore, in order to evaluate the reproducibility of the model, we compared the result of DNN with storage function model and Rainfall-Runoff-Inundation model (RRI) by some evaluation indexes. As a result, it is found that DNN model for the Shiokawa Dam can well reproduce and estimate stably the flood discharge relative to the other runoff models. Thus, it is suggested that DNN model is one of available dam inflow prediction method.

 We attempted to improve the accuracy of dam inflow forecasting using deep learning by augmenting the training data, and validated the applicability of the proposed method to multiple watersheds. The proposed data augmentation method which assumes a steady-state condition of constant rainfall, and uses a theoretical data set of virtual rainfall-runoff data as the augmentation data, such that the total rainfall and dam inflow into the watershed are equal (the runoff coefficient is 1.0). As a case study, the applicability of the data augmentation method to recent large-scale runoff was validated for four dam basins: Terauchi Dam, Miyagase Dam, Nomura Dam, and Kanayama Dam. The prediction accuracy of the data augmentation method was confirmed for each dam. However, when the test flood was much larger than the study flood in the past, there was a limit to the improvement in reproducibility.

 It is necessary to develop a river flow model with low numerical-parameter dependency for the evaluation of river flow variation due to climate change impact. In this study, we developed a river flow model with high accuracy and low computational load by coupling a 1D-2D hybrid numerical method and data-assimilation approach to evaluate the roughness coefficient. This model can keep low computational loads such as a 1D calculation model and high accuracy like a horizontal 2D flow model. The present model was applied to the flood flow in the Arakawa River under Typhoon Hagibis in 2019. The results indicate that the calculated roughness coefficient with the present model was reasonable compared to that with an existing 1D model. It is also noted that the CPU time in the present model was appreciably reduced to that in a general 2D calculation, demonstrating the fundamental validity of the present model.

 We constructed method for the damage risk estimation of the riverbanks and revetments. The method estimates the risk based on the followings: 1) the channel longitudinal and cross-sectional geometry. 2) the deepest riverbed level. 3) the stable riverbed level and width, and 4) the dynamic equilibrium longitudinal profile riverbed level. The method was then applied to the Hikosan River. The results showed the followings: 1) damaged riverbanks and revetments were included in sections where the river channel was not stable, and the river tended to widen, 2) The damaged riverbanks and revetments were included in the method's serious damage risk potential zone, and 3) The later flooding has been observed to have damaged riverbanks and revetments in the high potential section for damage based on the past channel geometry.

 The flow structures in the bend a meandering channel downstream of a negative step were investigated under the conditions with submerged jet and undular jump using laboratory experiments and three-dimensional (3D) numerical calculations. Comparison with experimental results in straight and meandering channels shows that 3D calculations using the standard k-ε model for Reynolds-Averaged Navier-Stokes (RANS) model can explain water depth distribution downstream from the step. The results of the crosssectional velocity distribution in the meandering channel downstream of the step show that the characteristics of the velocity distribution in a bend are different between a submerged jet and a undular jump. In particular, it was found that the secondary flow of the undular jump was enhanced rather than that of the submerged jet. It indicates that the strong secondary flow and accelerating bottom velocity increase the risk of riverbank erosion along the outer bank downstream of a negative step.

 This study summarizes the progress of the development of real-time inundation inside levee prediction method that aim at the evaluation of floor inundation. This method saves the depth distributions pre-calculated by dynamic models in preparation phase, and calls these depth distributions according to the arbitrary hyetograph in real-time operation phase. This reduces the computational load. At the time for verification of the proposed method, flood analysis results are compared between conventional two-dimensional indefinite flow calculation and proposal method. As a result, although the proposed method can calculate the time change of the inundation water volume, the proposed method cannot estimate inundation depth distribution at arbitrary time. However, if the proposed method calculates from the occurrence of inundation to the elimination, proposed method obtains maximum inundation depth equivalent to conventional. Therefore, the proposed method cannot evaluate the time of occurrence of inundation on the floor, but can evaluate the presence or absence of inundation on the floor.

 The purpose of this study was to develop an estimation method for the break-up period that occurs before ice jams occur. Ice sheet thickness decreases during the break-up period. Ice sheet thickness calculation considering snow cover, snowmelt, rainfall are highly accurate in their calculated values. The calculated minimum water temperature was found to be an important value. The acceleration of ice sheet thickness variation P_B, was used to estimate the break-up period. This method uses six input values, estimated the break-up period on the safe side. The ice sheet thickness when fully frozen is an important parameter too.

 Numerical analysis of a water accident at a low-head dam was conducted. In assuming a human-body falling to a overflow weir, posture change and movement of human-body were simulated by being combined with a solution of 3D multi-phase flow. From a series of simulations conducted with different combinations of upstream depth and downstream depth, it was found that “wave jump” regime has vertical circulational flow in addition to “submerged jump” regime, which has been regarded as danger. While a human-body is trapped in a long time in a circulation flow of “submerged jump” regime, trapping in “wave jump” regime tends to be shorter. Because a human-body is strongly shaken in “wave jump” regime, “wave jump” regime seems to be more danger than “submerged jump” regime. A human-body was modeled to have the bouyancy that is caused by a standard lifejacket. Simulation results show that while the wearing of lifejacket might contributes to the prevention of drowning at low-head dam, the buoyancy of a standard lifejacket might be insufficient to keep a human-body floating on a water surface in the cases trapped in a circulation flow of “wave jump” regime.

 The quantitative evaluation of the error due to the influence of the strainer inside the groundwater observation hole was examined by numerical analysis based on the Darcy-Brinkman equation. It was confirmed that the entire region can be analyzed in a unified and stable manner by treating the permeability at the boundary between the fluid region and the porous region by the harmonic mean method. In addition, in the case of a 13 % aperture ratio, it was quantitatively shown that it was 18 mm from the side wall as a region containing a ± 10% error with respect to the flow velocity at the center of the hole.

 Continuous embankment in the upstream valley plain generating many "closed floodplains" makes the natural retarding basin function of the floodplain decline. In this study, we investigated the characteristics of the closed floodplain in the Kitakami River, which has a long valley bottom plain of 85 km, and proposed a method of using them as a series of retarding basins that functions only for excessive floods. Numerical flow simulations were conducted to estimate the performance of retarding basins under the condition that the design flood of 150 years in return period attacks the river reach which is constructed to the level of 90 years in return period. The result showed the possibility that the flood peak discharge can be reduced by nearly 2-6 % by several a closed floodplain with a section length of about 5 km.

 Numerical flow simulation was conducted to study the performance of an inverse siphon to drain flood from the closed floodplain where the collapse of the Kinu River embankment caused a severe flood damage in 2015. The result showed that the siphon shortened the inundation period significantly. This numerical model was applied to design a flood retention basin composed by a side overflow dike at the upstream end and an inverse siphon at the downstream end of the closed floodplain. The result showed that the facility efficiently works to reduce the peak discharge of a flood exceeding the river channel capacity.

 In Chiba Prefecture, the yatsu topography is widely distributed, and using this topography is expected as one of the flood control measures. In this study, to construct a basin water routing model to evaluate the effect of flood mitigation in the Ichinomiya River basin, we did runoff analysis using hydrological topographic data with different grid resolutions and compared the calculation result of them. As a result, the peak discharge at 150m resolution was 64% of that at 30m resolution, and the calculated results were significantly different. It was shown that when the grid resolution is coarse, the local elevation difference peculiar to the yatsu is smoothed, underestimating the discharge flux, and as a result, underestimating the runoff to the river channel. The above showed that it is necessary to reflect topographical data with high resolution in numerical modeling in watersheds with yatsu topography.

 This study focused on the introduction of permeable pavements on roadways as one of countermeasures against floods being intensified by climate change. And analyzed their flood control effects. Previous studies, which is for small rainfall events in the small watersheds, have ignored the rise in the groundwater level during infiltration of rainwater into the ground, horizontal infiltration in the ground, or drainage of rainwater stored in the pavement. In this study, a runoff analysis method that takes these factors into account was developed to analyze the flood control effects of permeable pavements for large rainfall events in large watersheds. The analysis revealed that the introduction of permeable pavements can have a certain flood control effect on future rainfall, and that this effect depends on the groundwater level conditions.

 In this study, we estimated the effect of the introduction of a paddy field dam with different functions on the reduction of the runoff and inundation into the main river and characteristics of runoff control in each paddy field district in the Kuro River basin of the Shira River system in the Kumamoto Prefecture. The results showed that the introduction of the paddy field dam that covers 1687 ha of the paddy field produces the highest reduction in the runoff to the main river through a 50-mm outlet hole for the 10-, 30-, and 50-year probabilities of rainfall; it reduces the peak flow by 33.7 m³/s (5%) and inundation area by 54 ha（7.1%）for the 50-year probability of rainfall. Furthermore, the peak cutoff per hectare to the downstream of each paddy field district was estimated to be higher when the topographic gradient is larger and there is no backwater.

 Considering intense floods which have frequently occurred in recent years, flood control measure including flooding area called "River Basin Disaster Resilience and Sustainability by All" have been investigated. In order to investigate concrete flood control measures in medium and small river basin, it is important to establish a highly accurate prediction method for inundation areas and a quantitative evaluation method for the effects of flood control measures. In this study, the Karube river basin of the Takahashi river system was selected as the study site. The effect of flood control measures, which combines the suppression of pump drainage and the increase of storage by raising the ridges of rice fields, was investigated numerically using a river residential area combined inundation model. As a result, under the same rainfall conditions as the heavy rain in August 2021, it was clarified that when the pump drainage rate was reduced to 50% the inundation area could be reduced to the same extent by raising the ridge by 30 cm.

 In this study, an analytical model of the Yonebukuro drainage pumping station basin in Osaki City, Miyagi Pref, was constructed to evaluate the flood mitigation effects of Smart Paddy Field Dam, and inundation simulations were conducted with the conventional and smart paddy field dams under multiple rainfall scenarios. The results suggest that the smart paddy field dam is more effective in mitigating flooding than the conventional paddy field dam, since the smart paddy field dam does not over-flow the rice paddy even if all rainfall is stored when rainfall does not exceed the drainage standard water level of the smart paddy field dam. However, in the case of rainfall that exceeds the drainage base level, the flood mitigation effect is reduced when a conventional paddy field dam is compared with a smart one that always cuts the peak drainage outflow. This is because the outflow is the same as that in rice paddies without the paddy field dam when the rainfall exceeds the base level. In addition, it was shown that the loss of storage pockets when the rainfall peak shifts backwards leads to weakening of the flood mitigation effect.

 In order to grasp the potential flood prevention of the irrigation reservoir, the flood inundation model was performed for the whole of Japan assuming that the maximum reservoir capacity can work. In addition, the expected annual damage cost was calculated. As a result, the reservoirs has the potential to reduce the amount of the expected annual damage cost caused by floods throughout Japan by about 1.0 - 3.0%. The prefectures with the highest reduction effect for the expected annual damage cost were Kagawa, Hyogo, Nara, Hiroshima and Shiga prefecture, in that order. On the other hand, it was found that the Kanto districts have lower reduction effect. In some prefectures, the damage reduction rate varied as the return period of the flood.

 Torrential rain disasters occurred in Rokkaku River watershed in 2019 and 2021. Especially, in the left bank basin of midstream of the Rokkaku River and the Takeo River, the tributary of the Rokkaku River, enormous damage occurred by the inland water flooding. This area is vulnerable to the inland water flooding so that the riparian work so far in the river cannot solve the existing problem of the watershed. In this study, storage facilities of the watershed are investigated for the solution. Focusing on the fact that the wide farmlands spread in the lowland region of the study area and there are several major irrigation reservoirs in the mountainous region of the same area, flood control countermeasures considering the inland water regulation reservoirs and the irrigation reservoirs are studied using 1D and 2D analysis of the flood and inundation. As a result, there are effects to decrease the beginning inundation depth by the storage facilities against the same torrential rain in 2019. It is clarified that there are some differences between the location even in the same watershed.

 River bridges under flood flows have been sometimes washed away frequently due to the occurrence of excess flood and the capture of driftwood on the bridge girder. The purpose of this study is to clarify the effect of caputured driftwood on the fluid forces on the bridge girder by conducting a model experiment and a 3D flow simulation. The result shows that the fluid force on the bridge girder increased sharply even with a slight increase in the flow rate due to the captured driftwood. In addition, the larger the amount of captured driftwood, the greater the fluid force acting on the bridge girder, and the increase in the projected area contributes most to this. 3D flow simulation indicates that flow structure around the bridge girder with captured driftwood are mostly examined.

 In recent years, the damage of bridge runoff in flood flow has increased. Therefore, it is necessary to investigate its mechanism and to reveal its countermeasures. In order to reveal the mechanism of bridge runoff, this paper has been investigating a mechanism of a bridge runoff in flood flow during the 2020 concentrated heavy rain at Kumamoto in Japan, by using a 3-dimensional numerical simulation method taking account for the river shape, the river bed shape and bridge. It has been found from flood forces acting on the bridge obtained by the 3-dimensional simulation that a bridge runoff mechanism can be estimated.

 In recent years, river disasters caused by unusually heavy rainfall have occurred frequently in many areas, and the damage has been increased by driftwood discharged by mountain collapses. Most driftwood control measures are implemented as gully control. With the increase in the external force of heavy rainfall in the future, it is important for disaster prevention to efficiently remove driftwood in alluvial fan river channels as well. However, no clear driftwood control design standards or guidelines have been established for river areas. Therefore, in order to efficiently design the shape of the inflow section of driftwood trapping facilities installed in river bends, a driftwood tracking model was applied to curved channels with different curvatures, and the effectiveness of the driftwood trapping function was investigated.

 This study conducted numerical experiments using a high-resolution numerical simulator to investigate ocean physical structures around coastal structures with considering stratification. To evaluate the effect of coastal structure’s layouts on ocean physical structures, three layout types of cylindrical objects with diameter of 500 m were considered: (1) single cylinder (SC), (2) two cylinders parallel to the flow direction (TCsPl) and (3) two cylinders perpendicular to the flow direction (TCsPerp). The model was forced by steady flows and tidally oscillating flows with M₂ frequency. Under the steady flow conditions, vortex motions were higher for the two cylinders cases than those for the single cylinder case. The intensity of vorticity was not largely different between TCsPl and TCsPerp under the steady flow conditions. Under the oscillating flow conditions, vortex motions were highest for the TCsPl case. Vertical mixing was especially enhanced for the cases of TCsPerp under the both flows conditions. This study suggests that the number and layout of coastal structures influence on oceanic physical structures.