The outflow angles of the water jet from a transverse weir in a straight channel and a curved channel with a radius of curvature of 0.70m were investigated experimentally. More over, it was attempted ti cinsyract a simple formula for calculating the outflow angle . The main conclusions of this study are as follows; 1) The change in overflow angle with increasing or decreasing fluid number in both straight and curved channels is greater at the downstream end of the weir than at the upstream end of the weir. In the sub-critical flow conditions, the outflow angle at the downstream end of the weir depends not only on the fluid number, but also on in the inflow dischrages. 2) The effect of centrifugal force due to the radius of curvature on the outflow angle at the downstream end of the weir is observed under subcritical flow conditions. 3) The outflow angles estimated with the flow vectores were compared with the ones masured directly. Good agreements were obtained for the overflow angle at the upstream end of the weir, but not for the downstream end of the weir. 4) An experimental equation was proposed to calculate the outflow angle at the upstream end of the weir.

 In fluid dynamics, a method combining deep learning and governing equations (PINNs) has been proposed as a high-performance data assimilation technique. PINNs are expected to be effective for phenomena like flow on sandbar, where obtaining large data is challenging, but their applicability remains unverified. This study evaluated PINNs’ applicability to flow over alternating sandbars. Results showed PINNs’ predictions agreed with numerical results, except for high wave numbers, indicating their applicability to flow over sandbar.

 The causes of coastal erosion include a decrease in sediment supply from rivers and the reduction of river mouth terraces, which serve as sources of fine sediment. Understanding the sediment dynamics in coastal areas, including river mouths and river terraces, requires comprehending the complex interactions between sediment transport caused by flood flow and waves. As the first step in analyzing the dynamics of fine sediment in the coastal area, including river mouths, an analysis of flood flow and riverbed variation was conducted, integrating the downstream river sections with the surrounding coastal area. The initial shape of the estuarine sandbar, which significantly affects the morphological changes in the river mouth but is difficult to define accurately due to its considerable temporal variation, was a focal point of this analysis. Analyses of the initial shape of the river mouth sandbar were compared using direct observational data and shapes estimated to match observational data during and after the flood, investigating the process of estuary sandbar flushing and estuary terrace formation. The results show that the time change in water and the process of river mouth terrace formation vary significantly depending on the initial shape of the river mouth sandbar, highlighting the need to understand the shape of the river mouth sandbar before the flood.

 Recent advancements in remote sensing technology are enhancing the understanding of topography and particle size distribution. This study examines the effect of high-resolution topographical and particle size data from UAV on riverbed variation analysis accuracy. We conducted pre- and post-flood surveys using UAV photogrammetry and particle size investigations. Varying topographical resolution from 200m to 5m improved the average error in riverbed variation volume from 0.497m to 0.396m. Accuracy improved in areas with rapid longitudinal changes, such as sandbars. Detailed particle size distribution further reduced the error in riverbed variation volume.

 Seigyu is a kind of traditional spur dyke consisting of wooden frames and gabions. It alters flow velocities and directions, and has been installed to prevent river bank from erosion, to enhance the diversity of river environment and so on. In actual rivers, the jamming of driftwood at Seigyu is a common phenomenon during floods, which may affect the original function of the structure. In this study, experimental studies were carried out to investigate the effect of driftwood jamming on the bed deformation characteristics and sediment control functions of a Seigyu. The change of the permeability of the front surface of the Seigyu and the location of jammed driftwood were considered as two key factors. In the experiments, various permeabilities were tested for the former factor. While for the latter one, 4 kinds of jamming patterns were considered. Based on the experiments, the scour and deposition shapes and volumes, local bed profiles and the amount of Seigyu’s movement were analyzed. It was found that the jamming of driftwood led to an increase of bed deformations despite of permeability and jamming location. With the same permeability, the most significant bed deformations were found in the pattern jamming at the top of the Seigyu.

 This study developed a new non-hydrostatic, non-equilibrium, one-dimensional numerical model for bed deformation. The model used the depth-integral and moment equations with origin at half depth as governing equations. The non-equilibrium riverbed model could treat the non-equilibrium sediment discharge by the sediment layer based on the framework of the existing model. The model was validated against experimental data of dam-break flows and dike overflows on the movable bed. The verifications showed that the model could reproduce the experimental results of dam-break flow and overbank failure phenomena with sufficient accuracy. On the other hand, the stability of numerical analysis for the bed deformation and the pressure convergence remained an issue.

 It is well known that erosion due to abrasion occurs when sand and gravel in flow impinge on hydraulic structures composed of concrete. In order to evaluate the risk of damage to structures due to abrasion, it is important to have an evaluation method that can appropriately predict the abrasion phenomenon caused by sand and gravel. In this study, a prediction method of scour due to particle abrasion using CFD-DEM is developed, and its reproducibility was examined by applying it to experiments of scour due to particle abrasion on a concrete channel. The results showed that the scour depth increases in a larger fall height channel than in a smaller channel due to the increase in vertical particle velocity at the bottom of the slopes. The particle saltation reproduced the phenomenon of abrasion due to the collision of particles accelerated by the negative pressure at the slope shoulder. Since these are main features for the upstream migration process of a waterfall due to scour due to abrasion, the applicability of this calculation method to the prediction of abrasion scour is demonstrated.

 In this study, we propose a method using UAV-SfM that does not require a scale in images. We also addressed various image analysis issues through image processing and created an RPA (Robotic Process Automation) program for automatic continuous processing, reducing labor in measuring riverbed surface grain size distribution. We applied this technology underwater and found that for UAV images where underwater stones and gravel are distinguishable, image processing improved grain size identification accuracy. In high water depth areas where stones are not distinguishable, image processing slightly improved identification accuracy in shaded areas.

 We investigated the characteristics of changes in spectral reflectance with changes in turbidity using sediments collected near the estuaries of first-class rivers in the Shikoku and Kyushu regions, and confirmed that the spectral reflectance of turbid water depends on the type of sediment, as in previous studies. Furthermore, in order to explore the possibility of applying the turbidity estimation equation developed from the experiments to rivers with unknown turbidity, a cluster analysis was conducted using the experimental data and the spectral reflectance data of turbid water from the Sentinel-2 satellite. The results suggest that although it is possible to classify turbid water within a turbidity range of ±20% using only the experimental data, it is necessary to investigate the factors that affect the relationship between turbidity and reflectance. The Sentinel-2 satellite data did not reproduce the classification based on the experimental data, suggesting the need for further investigation of these factors.

 UAV-captured images and image analysis technology are being used to improve the sophistication and efficiency of river patrol operations, and the application of image generation AI to object detection models has recently been considered. To achieve highly accurate object detection, it is necessary to train a large number of images of various types. If generated images can be used to train object detection models for river patrol images, a large number of images can be acquired efficiently. In this study, we used the image generation AI “Stable Diffusion” for a plastic bottle, trained UAV images and generated images, and compared the detection accuracy to verify the appropriateness of applying the image generation AI to training data and whether the generated images can replace aerial photography images. The results suggest that it is possible to apply the generated images as training data, and that the generated images can be substituted for the training data if there are only a few real images.

 When applying commercial analysis software to estimate the shallow water bottoms using Unmanned Aerial Vehicle (UAV)- based photogrammetry, the depth of the water as shallow is often underestimated because the refraction of the line of sight at the water surface is not taken into account. This refraction causes significant errors in depth estimation, leading to inaccurate data that can affect various applications, including environmental monitoring and resource management. As a countermeasure, the estimated shallow water depth is corrected by multiplying it by an appropriate coefficient (Water Depth Underestimation Factor). However, some software programs do not disclose their internal triangulation (coordinate estimation) algorithms. Therefore, making it impossible to calculate the Water Depth Underestimation Factor in accordance with the algorithm. In this study, through the use of device to do triangulates underwater points using only photographs whose line of sight is close to the vertical, and conducted experiment for using actual aerial photographs of a small river to verify the effectiveness of this device. The results showed that the Water Depth Underestimation Factor, estimated without using the measured depths, was stable regardless of the 3D location of the underwater points when simple assumptions about the triangulation algorithm were made. Additionally, it was almost consistent with the optimized Water Depth Underestimation Factor derived using the measured depths. This demonstrates that the proposed method is capable of accurately correcting the depth of the shallow water.

 Analysis method of flood in wetland was studied using estimated elevation maps from satellite images and observed river water level. The study area is the haor region in the Meghna river basin located in the northeastern region in Bangladesh. Yearly changes in topography due to sediment storage or land use changes were analyzed with estimated elevation maps of different years. Flood inundation simulation was conducted using high-resolution elevation maps, then water depth time series was calculated. Effect of morphological changes on flood arrival time was estimated. This method applies to the layout design of water intake facilities to reduce the difference in water level between inside and outside of the embankment and mitigate damages to the embankment.

 In regions with insufficient meteorological observation networks, the provision of satellite precipitation data with a certain degree of quantitative accuracy at the watershed scale is expected to play a crucial role in disaster prevention, agriculture, and other socio-economic sectors. In this study, we proposed a new watershed precipitation estimation method using high-frequency and high-resolution observation data from the Himawari meteorological satellite and deep learning. We also quantitatively evaluated the estimated watershed precipitation across multiple rivers in regions where parameter optimization was not performed. As a result, for basin-averaged precipitation in watersheds larger than 100 km², we reproduced non-typhoon heavy rainfall with higher accuracy than GSMaP. Even in small watersheds with areas ranging from 0.2 km² to 100 km², the estimated precipitation demonstrated quantitative accuracy when compared to Thiessen precipitation calculated from ground rain gauges for non-typhoon heavy rainfall. These findings indicate that the proposed watershed precipitation estimation method possesses a certain level of quantitative accuracy regardless of watershed scale.

 In this study, we used a model that forecasts precipitation one hour ahead based on 2.5 hours of prior meteorological data using Convolutional LSTM. We examined the training data by adding weather elements other than precipitation and by applying data augmentation with different combinations of rotation, expansion, and movement rates. The results show that the Threat Score for precipitation greater than 50 mm h^-1 is 228% more accurate with the addition of mean sea-level barometric pressure and 217% more accurate with the application of data augmentation methods up to 30 degrees rotation, up to 4 times magnification, and up to 10% shift compared to the JMA's Very-Short-Range Forecasts of precipitation. The findings indicate that the accuracy of prediction of strong precipitation by deep learning, which had been considered an issue, could be improved by using appropriate training data.

 To improve the accuracy of typhoon forecasts, advancing data assimilation techniques is crucial to estimate the initial conditions of typhoon forecasts. In this study, we propose reconstructing Rankine vortices (RVs), which are used as an approximation model of typhoons, from Doppler wind data. We trained two deep learning methods, autoencoder (AE) and conditional variational autoencoder (CVAE), and succeeded in reconstructing RVs from noiseless and noisy Doppler wind data. While AE provides better estimates for RVs for noiseless data, CVAE outperforms AE for noisy Doppler wind data. Sensitivity experiments to latent-space variables reveal that VAE works as a generative model for RVs.

 Our study reports whether or not it is possible to make good predictions even in spatiotemporal computational domains where there are no training datasets using physics-informed neural networks (PINN), which are the deep neural networks (DNN) that extendedly incorporate physical laws. To predict flood events in drainage canal networks in low-lying agricultural land, the PINN that incorporates the Saint-Venant equations (i.e., PINN-SVE) was applied to a straight canal in the drainage networks. Owing to no in-situ observed data, simulated flood data generated by a rainfall-runoff model with pseudo heavy rainfalls were used as the correct values to predict flow velocity and water depth in canal cells without training datasets. The predicted PINN-SVE results using 100 patterns of flood events showed that the errors and peak differences from the correct values were generally better than those of the conventional DNN. Good prediction accuracy was obtained for waveforms with smoothly temporal changes in the flow velocity and water depth; however, the poor accuracy was likely caused by the waveforms with rapidly temporal changes.

 Since surveys of river space use do not capture the attributes of users and detailed usage conditions, it is desirable to develop a new method to complement the current survey results in order to properly evaluate the effects of river improvement and to understand the involvement of residents in rivers. Therefore, in this study, as a basic investigation of a survey method on people's use of river space using a river camera and AI together, we investigated the degree of recognition of attributes and behaviors of people using rivers by utilizing LLaVA, a type of LMM. In addition, we compared LLaVA with an object detection model to investigate the influence of the shooting environment, and examined the usefulness and application limitations of LLaVA. The results of LLaVA estimation of human gender, age, and behavior showed that men and women generally agreed with visual recognition, but elderly people tended to be misjudged. The results generally agreed with visual inspection except for some behaviors. The results show the usefulness of the model in recognizing human attributes and behaviors.

 Thermal Image Velocimetry (TIV) is a method for identifying the two-dimensional wind velocity field near the ground surface. In this study, we applied TIV in the detrection of spatial wind velocity distribution along the three-dimensional and complex surface of Mt. Fuji. This is accomplished by introducing the raytraicing technique. This works to project 2D thermal image in the camera to the 3D disgital surface model of the Mt.Fuji. Then, usual algorithm of TIV can detect the temperature fluttuaiotn pattern in the surface temperature could be tracked 3-dimensionally. As a result, spatial velocity distribution along the mountain surface was detected in a range of 4 km by 3 km of the southern side. The visualization of the movemet of the surface temperature pattern showed the local wind flowing around a local peak of Mt. Fuji. It is also visualized that the streaky temperature fluctuation pattern developed perppendiculat to the main wind direction of the ascending motion of along the mouain surface, which cannot be seen in an ordinal atmospheric boundary layer.

 The wind spectrum of the atmospheric boundary layer is composed of the two well-known energy peaks. These are, the high frequency peak due to convection in the boundary layer, and the low frequency peak due to baroclinic instability. While the high-frequency peak has been well studied in many previous studies, the low-frequency peak has not been examined as often. This is because long-term data is required to investigate low-frequency phenomena, and data collection is difficult. In this study, the characteristics and formulation of the inertial range that appears from 5 to 24 hours in the wind spectrum were examined. Wind data from meteorological observation towers were used to analyze wind spectrum for up to 17 years. The results show that the interannual variation of the wind speed spectrum is linked to the interannual variation trend of the annual mean wind speed, and that the energy dissipation rate depends on the mean wind speed.

 A comparison and validation of different sensible heat transport processes in three single-layer urban canopy models were conducted. The differences among them lie in their assumptions regarding the sensible heat transport processes between the various components of the urban canopy. The validation was performed using data from 17 geographically and climatically diverse cities that have been recently compiled. There was no single model that produced good results for all three fluxes: longwave radiation, sensible heat, and ground heat. The two models that aggregated sensible heat at the urban canyon particularly showed the impact of urban geometry. All three models produced similar phase changes to the observed data.

 We evaluated the effect of different sensible heat transport modeling on sea breeze fronts in urban canopy model, taking into account appropriate urban thermodynamic and aerodynamic effects. A total of 4 cases with different sensible heat transport models and different settings for the oasis effect of vegetation were calculated for sea breeze days in Kanto region in Japan. The results show that in the calculation cases with the Top-down method, which calculates sensible heat fluxes considering the coarseness and density of the entire urban canopy with and without the oasis effect, is used for the heat transport process, the sea breeze is delayed in the metropolitan area despite the strong pressure driving force between the land and sea surfaces. This is due to the increase in friction velocity, or turbulent mixing, with increasing roughness, indicating the importance of properly considering urban geometry and properly incorporating it into the modeling.

 This study proposed a method to instantly assess the spatial distribution of detailed turbulence statistics within an urban district by referring to an urban turbulence database calculated from LES simulation. The turbulence statistics within an urban district are generally determined by the urban geometry and the main wind direction of the mean flow. Based on this characteristic, LES calculations were performed for each of 16 representative wind directions, and a database of the spatial distribution of turbulence statistics in urban areas was created. The database is referred to in accordance with meteorological information from a wide-area meteorological station, and the representative velocity is scaled to enable instantaneous diagnosis of turbulence statistics within a urban district. The output diagnostic value of the mean wind speed has a linear relationship with the observed value, confirming that the mean flow within an urban district is generally determined only by the urban geometry and the main wind direction, thus demonstrating the validity of the method. On the other hand, when the accuracy of the representative wind speed values in the estimated area is insufficient, there is a difference between the diagnosed value and the observed value.

 This study uses machine learning to develop a model for estimating turbulence statistics in an urban area using building height, vegetation distribution, and elevation as input. Machine learning generally requires a large amount of training data. Therefore, we used a database of the turbulent statistics within an urban district which covers the entire Tokyo 23 wards in 2 m resolution. This database enables us to learn several thousands of 320 x 320 m² areas. The results show that the horizontal distributions of mean streanwie velocity, spanwise wind velocity, and turbulent kinetic energy at a height of 2 m above the ground can be estimated with significant accuracy. In addition, the spatial mean values of within the test area be estimated with higher accuracy. On the other hand, the accuracy of the estimation for open space with no building or vegetaion was not good, which may be due to the influence of the outside of the test area is more strongly reflected.

 In light of the recent emphasis on energy harvesting, the technology utilizing tiny amounts of energy around us to realize sustainable society, we focused on small wind generation that utilize energy from turbulent flows generated in urban areas. In this study, we developed an algorithm to estimate the instantaneous theoretical maximum power generation utilizing urban wind, which is highly variable with time and has large turbulence intensity. When the load resistance in the circuit was dynamically controlled based on the wind speed, it was found that the higher the turbulence intensity, the higher the power gen-eration efficiency under the same mean wind speed. We point out that dynamically controlled load resistance could improve power generation by at least 2.6 times compared to the fixed resistance case.

 The Lattice Boltzmann Method (LBM) is a fully explicit computational method with high parallel efficiency. While its development for tsunami analysis is active, there is only a few hydrodynamic models for environmental flows appeared in high aspect-ratio water bodies such as lakes and coastal area. Additionally, the use of the no-slip condition at boundaries increases computational cost as it requires a large number of lattices near the wall. Therefore, in this study, we have developed LBM employing high aspect-ratio lattice together with a wall function to reduce the computational cost in environmental flows. We examined the present LBM model with a lock-exchange flow and internal wave generation problems. Although the lattice flattering tends to cause the numerical diffusion near the density interface, the results show reasonable agreements with the previous numerical models and laboratory experiment.

 The accuracy of the diversion control of a high-accuracy diversion chamber, which can draw a certain flow rate from an unsteady open channel flow, was verified by computational fluid dynamics (CFD). First, we conducted a numerical simulation of an existing laboratory experiment to confirm the validity of the CFD. The accuracy of the diverted flow rate simulated by CFD was estimated to be equivalent to that by the experimental measurement. Next, we simulated 3 cases with different numbers and arrangements of orifice holes in the general design of a high-accuracy diversion chamber that is expected to be applied to a combined sewer system. In the case of a single orifice hole, it was difficult to control the diversion due to the effect of the jet flow from the orifice hole. On the other hand, in the case with multiple orifice holes, the effect of the jet flow was sufficiently mitigated. Besides, it was found that a high-accuracy diversion chamber allows for much more accurate diversion control compared to the conventional diversion chamber.

 High-resolution Shallow Water simulations have the issues of its computational cost and numerical instability due to the steep gradient of water surface. For these issues, hp-adaptive Discontinuous Galerkin (DG) Method has been proposed to control the grid resolution and the order of elements. However, the adequate adaptive strategies haven’t been well developed for shallow water problems, to decide how to control the grid resolution and order of accuracy against numerical oscillations especially for structured/unstructured mixed grids. In this study, we proposed a new adaptive strategy which combines the indicators of the cell average gradient and smoothness of the solution on the Systematic Grid which realizes the analysis on arbitrary grid. We applied the present adaptive strategy and the mixed grids to oblique hydraulic jump and equatorial Rossby modon problems. We confirmed from the numerical experiments that the proposed adaptive strategy effectively suppressed the numerical oscillation while keeping mass conservation and lead to efficient and accurate analysis on the mixed grid.

 Numerical computations that consider more detailed conditions are necessary to predict flood flow phenomenon. The Average Component Acceleration (ACA) method, developed by the authors, enables accelerated analysis by shortening the time scale of discharge hydrographs and adding averaged component terms to the time evolution equation for water depth. In this study, we clarified that the reason of the numerical oscillations appearing in the ACA methods with high acceleration rates is partly due to the riverbed gradient. Additionally, we successfully eliminate oscillations by weighting the averaged component terms. When the ACA method was applied to the GBVC3 method, it represented differences in water levels between left and right banks. Furthermore, the turbulence kinetic energy calculated from the analysis results closely matched that of the original analysis. Thus, the ACA method demonstrates the potential to accelerate analysis while maintaining the accuracy of quasi-three-dimensional analysis methods.

 This paper aims to develope a practical integrated analysis method of flood flows and waves. For this purpose, the quasi three-dimensional flow analysis method(Q3D-FEBS) has been extended by reconsidering the basic equation system and its numerical procedure to be applicable to wave propagation and spilling wave breaker (Q3D-FEBS-FWI).The applicability of the Q3D-FEBS-FWI was examined in experiments of spilling wave breaker and wave-current interactions. From these studies, it was demonstrated that Q3D-FEBS-FWI can explain the characteristics of wave propagation, as well as the phase-averaged vertical velocity distribution in the surf zone and wave-current coexistence field. Additionally, the challenges for developing an integrated analysis method for flood flows and waves were clarified.

 We have developed a SPH based Fluid-Structure Interaction (FSI) model to compute hydrodynamics around structures such as wave-dissipating blocks. To improve the accuracy of FSI simulation, we carefully investigate the neighboring search of the water and solid particles, labeling of solid particles, selection of kernel function. We also formulated and added frictional forces between solids required for designing structures composed of wave-dissipating blocks. The present model was validated with dam-break wave, beam vibrations, static friction, and the interaction between an elastic plate and fluid. The results are compared with previous research and analytical solutions, and showed reasonable accuracy in gravitational wave propagation, static frictional forces and the maximum displacement of elastic bodies.

 Althogh micro-plastic (MP) redispersion from coastal area has been pointed out, there is few study to quantify the MP dynamics and its interaction with sand under wave and currents. In this study, as a fundamental investigation for MP dynamics, we have developed CFD-DEM coupling model to understand interaction between solids as well as between solid and fluid. We developed a DEM solver and employed open-source software OpenFOAM for hydrodynamics, then coupled these models using preCICE library. We first introduce DEM formulation and preCICE library together with a coupling test. Next, we applied the present model to particle settling problems and found reasonable agreement with an analytical formula in 1-way coupling while discovering different behavior in 2-way coupling for single particle test cases. Lastly, in the multiple particle test cases, we observed changes in particle behavior due to the use of an empirical formula dependent on the porosity.

 A method was developed to analyze the scale of artificial fish nests that prevents filamentous algae, which detach from the canal bed and flow downstream (detached algae), from accumulating in rectangular artificial fish nests. Previously, it has been reported that a large amount of sediment, including detached algae, accumulating within fish nests can reduce the living space for fish, thereby impairing the effectiveness of the habitats. In this study, existing fish nests in the agricultural drainage canal in Ibaraki Prefecture were the focus, and numerical computations of three-dimensional flow and the three-dimensional behavior of cylindrical detached algae were conducted using iRIC NaysCUBE. Five different scales for the fish nests were set, and the "discharge rate," which is the percentage of detached algae that flows out of the nest due to water flow, was calculated. The results showed that when the entrance width and depth of the fish nest are reduced beyond a certain threshold from their current dimensions, the discharge rate of detached algae decreases.

 One of the conditions for a fish-friendly river is the provision of shelter, and one method of securing shelter for fish is the installation of fish nest blocks. Various studies have been conducted on fish nest blocks, and it has been found that fish nest blocks are useful as resting places for fish when they become fatigued and as shelter from external factors. In this study, the effects of aperture shapes (full open, upstreamopen, downstreamopen, top open, and bottom open) and velocity divided by body length (3, 6 and 9 (1/s)) on the resting characteristics of Plecoglossus altivelis, Opsariichthys platypus and Nipponocypris temminckii were examined. As a result, in the case where the aperture was secured on the bottomside, it was found that fishes can enter the fish nest block easy, and quickly. Spending time in the full open and bottom open fish nest blocks is longer than that of another one.

 Many experiments have been conducted to determine the optimum fishway standard on the basis of the migration rate of a single fish species released near the entrance of a fishway. However, fish try to migrate in the fishway surrounded by other fish species. In this study, the migration experiments in the pool-and-weir fishway were conducted as single or multiple species of Plecoglossus altivelis, Opsariichthys platypus and Misgurnus anguillicaudatus. In the single species condition, each species formed a school of fish at its preferred velocity and attempted to migrate upstream. In the multiple species condition, all fish species changed their schooling position as compares with that of single species. The school of P. altivelis moved to the low velocity zone, while O. platypus and M. anguillicaudatus were pushed away from the P. altivelis school. On the other hand, O. platypus are pushed into the high speed velocity area and become more motivated to migrate upstream.

 Alluvial fan floodplains are anthropogenically influenced aquatic ecosystems formed by river water withdrawals and groundwater. This study investigates their thermal water environment and changes in fish assemblage from spring to autumn using environmental DNA metabarcoding. The results showed a remarkable difference in temperature between the main river and floodplain water bodies. In summer, the flood-plain channel receives warm water, but areas influenced by groundwater discharge remain low and stable in temperature. Fish appearance characteristics varied, with species disappearing from the main river and others appearing only in the floodplain water bodies during the summer. No clear characteristics were identified, but these floodplains may provide valuable cool habitats as global warming continues.

 This study aimed to improve the detection efficiency of environmental DNA analysis of gobiid fishes inhabiting tidal flats in brackish water areas of rivers. The results of environmental DNA analysis using pore water were compared with those of a sampling survey. The comparison results suggest that environmental DNA analysis using pore water is more efficient than collection on gravel flats, whereas environmental DNA analysis of pore water is less efficient than collection on sand and mudflats, especially on sandy flats.

 In recent years, many ecosystem-friendly revetments have been installed in small and medium-sized rivers as Nature-oriented river works. One example of such revetment is a frame revetment that is porous to provide habitats for aquatic organisms. A salt marsh is also important as a habitat in an estuary. In this study, we conducted an environmental DNA survey of Japanese eels and observed the frame revetment using an underwater camera to observe fish behavior.

 The results showed that fishes such as Japanese eels and Chichibu goby were using the frame revetment as a resting and feeding site. The environmental DNA analysis results showed that the environmental DNA of the Japanese eel at the bottom of salt marsh was not detected during hypoxia, and it was detected before and after hypoxia. On the other hand, environmental DNA concentration at the frame revetment and surface increased during hypoxia. This suggests that Japanese eels temporarily changed their habitat from the bottom of salt marsh to the surface of it and estuary frame revetments in response to the hypoxia caused by rainfall and tidal changes.

 In flood evacuation using My Timeline, residents themselves determine in advance the criteria for starting evacuation (when evacuation should begin). However, a systematic method for determining the criteria personalized for individual residents has not yet been developed. In this study, under conditions where river water level is used as evacuation trigger, we calculated the maximum water level at which evacuation succeeds (maximum evacuable water level) for each departure point using five rainfall scenarios with different return periods on the floodplain of Serikawa River in Shiga Prefecture. The results show that the relationship between magnitude of rainfall and maximum evacuable water level varies depending on the floodplain type. In area affected by main river flooding, maximum evacuable water level is lower for the 100-year and 200-year rainfall event. Conversely, in area affected by only inland flooding, maximum evacuable water level is lower for the 10-year rainfall event. These differences are attributed to factors such as rainfall patterns and land use/microtopography around the departure points, which can significantly impact the critical evacuation water level.

 The heavy rain disaster occurred at Kawanishi-hatakura, Yufu, Ono and Otsuru, Hita, and Miyazono, Nakatsu cities in Oita Prefecture from June 30 to July 10, 2023. To investigate the correlation between past evacuation experiments and this disaster evacuation behavior, and the reason of this evacuation behavior, we interviewed the residents of four communities. We identified the correlation between past evacuation experiments and this disaster evacuation behavior. The experience of past evacuation had the effect on the resident evacuation behavior by this disaster. The reason of the evacuation was an evacuation call and understanding of danger around the resident. The reason of the non-evacuation was the evacuation call activity, normalcy bias, lack of understanding of danger around the resident, and the selection to stay in the home. Residents with past evacuation experiments thought that the distance to the evacuation shelter was more dangerous than staying at home.

 The effectiveness of the information transmission is important to be investigated for the safe evacuation against intense and frequent flooding damage. In this study, a method of evacuation analysis has been developed that takes into account changes in evacuation behavior due to information sharing among evacuees and transmission of inundation information at evacuation centers. By conducting evacuation analyses for 15 levee breaching locations in Kawashima Town, Saitama Prefecture, the effects of information transmission and the addition of evacuation centers on evacuation characteristics were investigated. Information sharing among evacuees is effective in shortening the evacuation distance except under some cases, and sharing information on flooding at evacuation centers is effective in reducing the risk of encountering flooding. Both of these increase the success rate of evacuation, especially for residents with the nearest evacuation center near the levee breaching point. Although closed shelters increase the travel burden of evacuees, the success rate of evacuation can increase when shelters near the levee breaching point are closed. This suggests the importance of providing information on the exact levee breaching location.

 One of the countermeasures against excessive floods is a bulwark, which is a part of a river that is widened to a large extent and has an overpass levee installed. Hydraulic model experiments were conducted on this dam, and it was confirmed that it generates horizontal circulation flow and prevents the outflow of sediment into the dam. However, due to the limited number of measurement points, it has not been possible to elucidate the detailed flow conditions within the Tongtang. Therefore, we performed calculations to reproduce the model experiment using numerical calculations, and as a result, we were able to reproduce the model experiment at the water level. Using these results, we investigated the hydraulic characteristics of the tutang and found that the overflow from the overpass levee reduces the water level upstream and downstream, and the wide high water bed of the tuttong has a negative impact on farmland. It was found that there is a function to deposit.

 Because the frequency of major floods has increased in recent years due to climate change, it is considered necessary to implement flood control measures that anticipate a certain degree of overflow flooding. In such a situation, traditional flood management are expected to be helpful, but sufficient functional evaluation is required before they can be applied to actual rivers. The purpose of this study is to clarify the effect of the overflow levee on the flow in Kutsuwa Domo. We conducted numerical simulations with different locations and heights of the overflow levee. As a result, the following findings were obtained. The downstream overflow levee drains water entering through the low waterway. The upstream overflow levee drains water entering through the high water bed. With the low overflow levee, the overflow increased, and the backflow occurs on the high water bed.

 By concerns about an increase in excess floods due to climate change, there is a growing need to increase the flood detention function in river upstream area. On the other hand, closed floodplains, which are common in the upper reaches, are mainly used as rice paddies, such lands possibly to be used as temporary "running water type retarding basins" during excess floods. In this study, we designed a running water type retarding basin at a closed floodplain in the upper reaches of the Omono River in Akita Prefecture by assuming that a tentative levee development plan with a flood flow rate of 2800 m³/s has been completed. A numerical simulation about its performance of retarding basin against an excess flood of 4400 m³/s showed that it is possible to reduce the peak flow rate of the excess flood by approximately 12%.