The characterization of geomaterials in the near-surface plays a crucial role in geotechnical and geophysical engineering. Surface wave methods, utilizing Rayleigh and Love waves, are widely employed for near-surface characterization due to their broad exploration scale. However, the inversion process in geophysics faces several challenges due to the ill-posed nature of the problems. In this paper, we propose a data assimilation approach based on the ensemble Kalman filter (EnKF) to estimate the parameters of geomaterials using active Rayleigh waves data. The EnKF is a popular ensemble data assimilation method known for its ability to handle high-dimensional and nonlinear problems. It approximates the distribution of state variables using a ensemble of a small number of samples. Unlike gradient-based generalized linear inversion methods, the EnKF-based inversion does not require the calculation of a Jacobian matrix, and the parameters are estimated based on the statistical relationship between parameters and observations. The proposed method enables joint inversion, allowing the simultaneous inversion of different types of observations to reduce non-uniqueness in the inversion problem. Numerical experiments demonstrate the effectiveness of the EnKF-based inversion method in estimating one-dimensional soil property models using dispersion curves as observation data. The method achieves accurate results with a small number of samples and provides quantitative estimates of the uncertainty. Joint inversion of dispersion curve and first arrival data was employed to estimate the parameters of a two-dimensional model. The results indicate that the accuracy of joint inversion using multiple types of observed data is higher.

Rockfall movements are characterized by 3 key indicators: runout distance, M_l, maximum kinetic energy, M_e, and maximum jumping height, M_h. This study investigates the influence of material property variations on rockfall movements and proposes a highly accurate prediction model based on the traditional Response Surface Methodology (RSM). Five material properties are considered, they are young’s modulus for rock, E_rock , density for rock, ρ_rock , young’s modulus for slope, E_slope , coefficient of restitution between rock and slope, COR, and kinetic friction coefficient between rock and slope, μ_f . When comparing rockfall movements without material property variations to those with material property variations, it is observed varying degrees of difference. That is, the mean value of difference shows little influence but the standard deviation (SD) of the difference of runout distance varies from 9% to 35%, maximum kinetic energy varies from 0% to 2% and maximum jumping height varies from 8% to 86%. Additionally, a developed approach based on the traditional RSM called ‘MCS+RSM’ is proposed by using fewer input material parameters to reach a higher fitting accuracy. Specifically, compared to the 5 input material parameters in the traditional RSM, the prediction using MCS+RSM only requires 1 or 3 parameters. For the same number of sample points as the traditional RSM, MCS+RSM can accurately predict the mean and the SD value of runout distance, M_l, and maximum jumping height, M_h, while the traditional RSM has a large error in the prediction of SD value.

Landslide and slope failure are a continuous geological process, which should be analyzed through the whole process from pre-failure to post-failure. The transition of slope from stable to unstable conditions involves a major distortion of the soil and a change in constitutive behavior, where the post-peak strength degradation governs the response. Strain softening after soil peak strength is the point of concern through-out the whole analysis process, and its impact deserves to be discussed. In this study, the material point method (MPM) is used to analyze the whole process of slope failure with the strain softening elasto-plastic constitutive law. First, we verify the accuracy and necessity of considering strain softening for the slope stability stage and post-failure stage. In addition, the effects of softening parameters and soil spatial variability on the post-failure process are discussed. The main conclusions are as follows: (1) The strain softening behavior not only affect the slope stability, but also affect the runout characteristics of post-failure slope. (2) The softening stiffness S has no influence on the runout characteristics. (3) The MPM analysis for slope with soil strength spatial variability shows that the distribution of runout characteristics becomes wide due to strain softening.

Hierarchical Cartesian grids are a practical approach to conducting an efficient flow simulation of complex geometries. These grids are also helpful in simulating a free-surface flow in horizontally broad rivers and flooded areas with complex terrains and hydraulic structures. This paper proposes using block-based hierarchical Cartesian grids for free-surface flow analysis using one-phase volume-of-fluid approach. Special treatments at a refinement interface between adjacent blocks near the free surface, which are related to the liquid-flux calculation and the pressure interpolation, are presented that should be considered in the free-surface analysis with the hierarchical Cartesian grids. The results of two-dimensional numerical tests for flows with wave and advection show that the present numerical method can simulate the free-surface flows correctly the same as regular grids with no refinement interfaces.

This study aims to refine a three-dimensional flow model in a generalized curvilinear coordinate system by incorporating a porous media approach to simulate simultaneously both surface and seepage flows. The accuracy and effectiveness of the model are verified through its application to three scenarios: a dam break process over permeable porous bed, open channel flows with bed suction and curved open channel flows over a deformed riverbed. The numerical results demonstrate that the model is capable of accurately predicting water surface and velocity distributions over the riverbed with the complex boundary.

A neural network is required to be coordinate system independent when it is applied to solve physical problems. Instead of training with augmented data, this paper develops a post-processing which is applicable for any arbitrary neural network. The essence of the post-processing is synthesized rotation that generates a set of rotated input data, applies the neural network to each of the set, re-rotates the output, and extracts a coordinate system independent solution from the output. Numerical experiments of the post-processing are carried for structural response analysis. It is shown that the post-processing works as designed. Discussions are made for further use of the post-processing of neural networks to make them more physically admissible.

 In basins where disastrous floods are driven by long-duration processes, such as snow accumulation/melt and series of storm events, estimating maximum precipitation (MP) for not a single storm duration but long durations is necessary to estimate maximum flood. Although the model-based methodology to estimate MP for long durations was recently proposed, it is still required to better understand what characteristics of water years lead to higher increases in precipitation depths. To examine this issue, this study performed the model-based maximization of precipitation sequences during winter (Oct-Mar) of the 9 historical water years (94 Atmospheric River (AR) events) for the drainage areas of Bonneville Dam and Libby Dam in the Columbia River Basin. The storm selection based on the thresholds of Integrated Water Vapor Transport (IVT) showed that the use of CFSR resulted in selecting more AR events than the use of 20CRv2c-ensemble mean, implying the importance of reanalysis products in MP estimation. The 1996 water year showed high potential in precipitaton increase, implying that the 1996 Pacific Northwest floods could have been more disastrous. The total duration of AR events during winter in a water year showed high positive correlations (R=0.84 for Bonneville and R=0.97 for Libby) with the precipitation increase rate from historical to maximum in the long-duration precipitation depths. This study found that each AR’s trajectory may be the key to determining the increase rate in winter cumulative precipitation depths. Further studies are needed to examine how the use of different reanalysis products can affect the results and to quantify the potential in precipitation increase using AR trajectories.

 Because of recent severe flooding events, numerous embankments have been breached, resulting in large amounts of floating litter in residential areas. Understanding the behavior of floating litter during flood events is crucially important. For this study, we improved the comprehensive flood analysis model to reflect drainage from sluiceways and pumps. Moreover, we considered a detailed flood analysis until the floodwaters of the urban area were drained. Furthermore, a simulation of floating litter movement was performed using flood analysis results. Results show that more than half of the floating litter generated in the urban area during flooding remains in the residential area, but about one-fifth of the litter flows out through the breached site and flows into the river. The proportion of litter flowing out is higher in areas closer to the breach locations. Additionally, around 13% of the litter accumulates at drainage gates.

 Effects of past human activities and recent climate change disasters and their interaction on the riverbed morphology of the Chikugo River estuary, Japan were investigated using the long-term measurement of annual and seasonal topographic surveys. The results showed that human activities caused the extensive riverbed incision of the Chikugo River estuary with a maximum value of about 4 m from 1953 to 1998. After stopping the sediment extraction in 1999, the elevation of the estuary increased by 1.2 m due to the sediments transported by the tide, especially during 1998-2003. Starting from 2009, the occurrence of floods and landslide disasters increased in the upstream of the Chikugo River. Although disasters carried huge sediments, a decrease in elevation was observed between 10-17 km with a value of 1.5 m, which means that sediments supplied by disasters had been deposited in the upstream where sediments were extracted in the past, thus, they couldn’t be able to reach the estuary yet. So increased river flow with insufficient sediments eroded not only the surface silt and clay layer but also the bottom sand layer, resulting in a drastic decrease in the elevation of the estuary. Therefore, this study concludes that the riverbed morphol- ogy of the Chikugo River estuary is affected by both tidally induced sediment transport from downstream and an increase in river flow by recent disasters and human disturbances that were implemented in the past.

 Groynes installation is a popular measure to prevent riverbank erosion in braided rivers. In this study, evaluation of groynes installation has been done at a vulnerable reach of the Jamuna River in Bangladesh. Four groynes (officially termed as cross-bars) have been installed at the erosion prone (right) bank of the river adjacent to Sirajganj town. Due to the unpredictable behavior of the braided river and critical climatic conditions, stability assessment of the groynes is crucial, as well as their impact on surrounding area. To perform the evaluation, 50 km reach of the Jamuna River has been analyzed by numerical model using iRIC Nays2DH software. After successful calibration and verification of the model, the simulations have been done using peak discharges for six different cases. Analysis of the model output reveals that cross-bar 3 is in vulnerable condition among the four based on scouring at the toe of the groyne. Additionally, the impact includes formation of a weak zone at 4.0 km upstream of the cross-bar 1. More than 2.0 km of riverbank becomes susceptible to erosion due to the attack of the stream flow, which shows the impact of the constructed cross-bars in the surrounding area. This study provides insights into the stability assessment of groynes and their impact in braided rivers.

 Stable water isotopes (H216O, H218O and HD16O) are proxies of climate processes occurring in the water cycle, the latter having a crucial role in the Earth’s climate system. For example, water isotopes measured in polar ice cores and Asian speleothems are commonly used to reconstruct past temperature changes in the poles or to study the past dynamics of monsoons, respectively. In this study, we present the ongoing development and first results, for the preindustrial period, of a newly developed isotope-enabled version of the fully coupled Earth System Model MIROC6, called hereafter MIROC6-iso. Our preliminary results are in rather good agreement with isotope observations in precipitation and surface sea water. Some improvements are expected in Antarctica thanks to the coming implementation of water isotopes in the sea ice module of MIROC6-iso. The modeled isotope distribution in the first 1500 meters of the ocean goes also in the right direction compared to observations. For the deeper part, more spin-up time is necessary to reach a quasi-equilibrium state.

 Considering the applications using isotope-enabled fully coupled climate models from Europe and the United States for paleoclimate reconstruction, there is a pressing need for Japan to develop such a state-of-the-art model, too. Additionally, it paves the way for the future refinement and development of MIROC6-iso, while also serving as an important point of reference for subsequent analysis of the climate, hydrological cycle, and paleoclimate reconstruction using MIROC6-iso.

 In the statistical prediction of typhoon-induced rainfall, Fuzzy C Means (FCM) has been widely utilized to identify similar typhoons based on track positions (latitude and longitude). However, despite the potential impact of additional variables on rainfall distribution (e.g., low central pressure and slow movement speed resulting in high mean rainfall), their inclusion in similar typhoon identification using FCM has been largely unexplored. This is particularly relevant given the increasing strength and slowing down of typhoons due to climate change. To address this gap, we conducted a study incorporating these additional variables alongside track positions and evaluated their performance using various statistical measures. Our findings indicate that including central pressure and/or movement speed has little to no impact on the accuracy of rainfall prediction. The statistical measure values do not directly point to the direction of improvement or degradation of accuracy. As we confront the challenges of climate change, new and improved techniques will be necessary for accurate rainfall predictions.

 The Weakly Compressible Smoothed Particle Hydrodynamics (WCSPH) has been applied to simulate the flow through a vertical Gravitational Vortex Turbine (GVT) configuration and the motion of the turbine of a small hydropower system. The turbine blades are assumed to be rigid bodies moving with one-degree of freedom, which is the rotation about its shaft due to the hydrodynamic forces of the flow. The intake flow is an open channel flow which falls into a cylindrical basin which houses the turbine and impinges on the turbine blades. The method of simulation is the Weakly Compressible Smoothed Hydrodynamics (WCSPH) with the turbulence models in the equations of motion and the boundary conditions. The results are compared with a laboratory experiments conducted for the same dimension and the configuration. The flow features agree well and the motion of the turbune with a representative mass and a load agrees reasonably with the laboratory experiments.

 The destruction of trees in the river (riparian trees) has been mainly analyzed with consideration of flood loading recently, and the decreasing of the critical resistive moment used to evaluate the destruction condition was not analyzed in those studies though the actual resistive moment may decrease with substrate scouring by the heavy flood. In the context of climate change, which increases the likelihood of simultaneous floods and storms, the mechanism of riparian tree destruction under both flood and wind load and the change of critical resistive moment with riverbed erosion should be investigated. We established a mechanical model and analyzed the criteria for the destruction of riparian trees under flood and wind loads. One simple model was also proposed with considerations of root-soil plate radius and erosion depth of river bed to analyze the decreasing of the critical resistive moment. The taller and smaller trees are more vulnerable to be destructed than medium trees, with the consideration of both flood and storm. The critical resistive moment decreased as compared to the previous study, and the usefulness of the established model was verified to some extent.

 The effects of three-dimensional flow in strongly curve channel with partly emergent vegetation were analyzed numerically by comparing the flow structure of the Bottom Velocity Calculation (BVC) method and conventional two-dimensional calculation (2DC) methods. Numerical calculations result show that BVC method is able to reproduce the flow structure investigated by the experimental dataset, decelerating the surface velocity and accelerating bottom velocity. The presence of vegetation on the inner bank reduces the strength of the secondary flow, while the outer bank vegetation increases the strength of the secondary flow. The secondary flow reduces the streamwise velocity in the non-vegetation region and moves the maximum velocity toward the outer bank. These results show the importance of considering the 3D flow effect for simulating partly vegetation in meandering rivers.

 Soil moisture (SM) is a key component in land surface model (LSM). An accurate prediction of SM is therefore important since it affects other variables when estimating water and energy budgets. This study aimed to evaluate the impacts of SM schemes in a simple biosphere model including urban canopy (SiBUC) on water balance and streamflow estimations. SiBUC calculates SM of a three-layers soil model using Richards’ Equation (RE) with an explicit-midpoint method. The default soil parameterizations are based on the Clapp and Hornberger equation, and the van Genuchten (vG) function, popularly used in hydrological models, was newly added here. In SiBUC, the thickness of the first soil layer was 2 cm, whereas the second was several meters thick. Because of the RE’s nonlinear behavior, this coarse partition may not accurately estimate SM because the changes of a hydraulic gradient in near-surface soil may be large, requiring a finer discretization. A reference model called HYDRUS was used to assess SM by SiBUC. For sandy loam and loam soils, the SM by the default schemes in SiBUC was close to that predicted by HYDRUS but overpredicted for clay loam. Applying a finer partition in the near-surface soil and an implicit-midpoint method seem to improve the estimated SM and reduce the overestimated discharge in the Ping River Basin in Thailand. However, because of the relatively large time steps applied in SiBUC, the estimated runoff using vG was less accurate, possibly causing an overestimate of streamflow. Overall, the potential problem of SM using the RE was explored here, which may help the LSMs users to alleviate the numerical errors resulting from the inadequate discretization of the soil layers.

 In 2022, intense heat waves occur all around the world. In this study, we project these heat waves into the future with futuristic projection of hourly varying spatially distributed anthropogenic heat for three megacities—Delhi, London, and Tokyo. Different future climate forcing (CMIP5 and CMIP6) was also compared. We found that if similar heat waves occur in the future, they may be 0.8 to 1.5 °C hotter than the past events, on average. Urbanization in Delhi may severely worsen the heat wave, while projected decrease in energy usage in London and Tokyo may make the heat waves less severe. For the concerned heat wave events, urbanization effect was also found to be stronger in nighttime than daytime and exhibits large spatial heterogeneity and dependence on background climate forcing. Difference between CMIP5 and CMIP6 was significant but was much less than the difference between CMIP5/CMIP6 and the present.

 Flood forecasting is one of the most challenging aspects in the field of hydrology. With new developments in computational intelligence, data-driven methods are gradually becoming popular and Long Short-Term Memory (LSTM) approach has shown great potential in accurate river stage forecasting due to its ability to learn long term dependencies. This study investigated methods for further improving accuracy of flood forecasts provided by LSTM models, especially when the prediction lead time is increased. LSTM model simulations were carried out for data obtained from nine river basins in Sri Lanka and forecasts made from 1 hour to 24 hours lead times were analyzed. Different scenarios were used to train the model and the results indicated that using selected data from separate river basins has improved the forecasting accuracy, significantly for longer lead times. A better feature selection method (to be used as input data to train the model) was investigated in this study by evaluating the strength and relationship of the river water level variation among different river gauging stations. Using this feature selection method for selecting optimum water level data combined with rainfall data provided the highest accuracy in the predictions.

 This paper presents a study on the performance of the ensemble optimal interpolation (EnOI) scheme with stationary covariance matrices for assimilation of synthetic water level observations into the rainfall-runoff-inundation (RRI) model. Five state and observation error covariance matrices are picked from the ensemble Kalman filter (EnKF) simulation of a flood event and then used during the update stage for Kalman gain calculation for (i) state and (ii) parameter estimation with the EnOI scheme. The stationary EnOI improves state estimation at randomly selected validation locations (along with most of the river grids) for all five covariance matrices, but improvement throughout the basin is not guaranteed as performance is degraded at some unobserved locations. The method also has the potential to be used in real applications as the results improve not only for the same flood, but also for a different event from which the matrices are not extracted. Model parameters can also be successfully estimated with the EnOI scheme, but the estimation may be compromised if the parameters interfere with each other. Independently updating, or only updating the most sensitive parameter (Manning’s roughness coefficient for river in this case), leads to better estimation.

 After Minamata disease occurred in Kumamoto Prefecture, the Minamata Bay Pollution Prevention Project on Minamata Bay area was started to deal with mercury pollution. Beneficial from the project, the mercury content in the bay area has been significantly reduced. However, many studies have shown that the residual trace mercury content around the bay is still high. Furthermore, mercury has migrated from Minamata Bay to the Yatsushiro Sea. Therefore, continuous research on the distribution and content of mercury in this region is required.

 In this study, using sediment classification, the relationship between sediment particle size, particle specific surface area, and T-Hg (total mercury) concentration was investigated. Based on the numerical simulation, the particle size effect on the migration of mercury-containing sediments and the T-Hg distribution in the Yatsushiro Sea were studied. The result showed that the larger the particle size, the lower the T-Hg concentration in the sediment. Meanwhile, the larger the particle specific surface area, the higher the T-Hg concentration. Moreover, the smaller the particle size of the mercury-containing sediment, the higher the migration speed as well as the more comprehensive migration range. From the numerical simulation results, it was found that mercury was mainly distributed in the southwest and northeast directions from Minamata Bay. Finally, the simulated mercury distribution, which considers the relationship between particle size and mercury concentration, shows a high agreement with the past measurements of mercury, suggesting the importance of particle size effect in mercury migration.

 The impact of topography and tidal forcing on counter-current flow formation in monsoon-affected multi-channel macrotidal river estuaries is seldom studied. This research investigates the occurrence of counter-current flow during the monsoon season, characterized by strong semi-diurnal tidal currents, at the confluence area of the complex multi-channel macrotidal Tanintharyi River estuary (TRE) in Myanmar. Continuous measurements of velocity and discharge, along with surveys of topography and water quality (salinity, turbidity), were conducted during a 12-hour intensive survey (30-minute intervals) at the upstream junction area of the main channel and the branch during the spring tide. The main channel’s width, approximately four to five times larger than the branch’s width, experiences counter-current flow during ebb tide when water levels between the two channels are similar. Despite the freshwater environment and stable turbidity indicated by vertical profiles, tidal effects still influence counter-current flow formation. The imbalance in area and discharge ratios between the main channel and branch can be attributed to another important factor: downstream topography. The main channel with wide (1.3 km), straight (SI: 1.1), and short (28 km) topography, is highly sensitive to tides compared to branch with the narrow (180 m), meandering (SI: 2.1), and long (40 km) topography. The difference in downstream topography between the main channel and branch, combined with tidal forcing, plays a crucial role in counter-current flow formation in the upstream junction area of the TRE.

 Historical land cover data is crucial for understanding urbanization dynamics, climate modeling, and monitoring water resources. Following recent advancements in deep learning for processing Landsat archive data, prior studies have released high-resolution historical land cover maps on a global scale. However, these works often present prediction results limited to specific periods of coverage, which hinders their utility in conducting time series analysis across different urban agglomerations. To address this issue, we propose deep-learning models for land cover classification from Landsat images at a 30-meter spatial resolution. Our models are specifically designed for urban areas and are trained to be compatible with the sensors used in the Landsat series from 1972 to the present. Experimental results demonstrate that our models are highly effective in predicting land cover maps in new cities, particularly in built-up land and water regions. Our research provides pretrained models for land cover classification, facilitating future studies in related fields.

 This study presents a downscaling method that utilizes a machine learning algorithm to improve the regional analysis capabilities of general circulation models (GCMs) in hydrological prediction. The U-Net algorithm, a representative approach based on Convolutional Neural Networks (CNNs), is employed to emulate the dynamic downscaling process of regional circulation models (RCMs). In this process, the GCM output is utilized as the input, while the output from the regional climate model (RCM) acts as the label data. The accuracy of the model is evaluated by considering various factors, including the input datasets, model structures, and output formats. The input datasets consist of essential atmospheric variables such as precipitation, temperature, and humidity, which are examined in both two-dimensional and three-dimensional formats. The model structure is assessed by testing different hyperparameters and analyzing their impact on the accuracy of the predictions. Furthermore, the effect of the label data is also investigated. The results of the experiments demonstrate that the U-Net algorithm successfully emulates the dynamic downscaling process, even in the absence of precipitation input data. Increasing the number of convolutions in the model structure contributes to improved accuracy; however, enlarging the receptive filter size does not necessarily yield better results. Moreover, incorporating temperature data along with precipitation in the labeled dataset enhances the accuracy of the predictions compared to using precipitation alone.

 In this study, as a first attempt, we utilize the method of Radial-Basis Function Network (RBFN) to examine hourly rainfall intensity and soil-water index obtained from climate projections for investigating features of future rainfall under climate change influences. Our study area is focused on Rokko Mountain in Hyogo prefecture. The 1-km MLIT 3^rd-mesh system is adopted in our analysis. The 2-km and 5-km Non-Hydrostatic Regional Climate Models (NHRCM) under RCP2.6 and RCP8.5 scenarios are used as future climate projections. A transformation matrix is used to quantify geometrical changes of RBFN contours. The characteristics of cumulative rainfall amount and rainfall duration of hazardous rainfall events are also examined. The analysis results show that soil-water index is much enlarged while hourly rainfall is a bit enlarged under RCP2.6, but, reversely, soil-water index remains unchanged while hourly rainfall is much intensified on the mountain side facing Osaka Bay under RCP8.5 in 2-km NHRCM. The hazardous rainfall events in 2-km NHRCM have shorter durations and smaller cumulative amount comparing to the ones in 5-km NHRCM.

 Developing a quantitative flood damage estimation method is important to plan and implement effective adaptation and preventive measures for reducing flood damage to residential areas. We thus developed a flood damage assessment method by integrating hydrologic-hydraulic simulation model outputs and a flood damage model capable of considering house characteristics, and conducted the quantitative assessment of flood damage to houses and household contents, including the effectiveness analysis of flood damage reduction by the use of existing dams for flood control and elevating the plinth level of houses. We also developed average flood damage curves as a function of flood depth for the houses considering house types and also for household contents, using secondary data sources. Selecting the Solo River basin of Indonesia as the study area, we computed flood characteristics using a water and energy budget-based rainfall-runoff-inundation model and estimated flood damage using the developed flood damage curves based on house characteristics and the exposed property value. The findings show that the use of the dam for flood control in the basin can reduce the expected annual damage (i.e., the average total expected damage per year from all flood probabilities) to residential areas by 21%. Flood damage to households can be reduced by more than 45% by elevating the plinth level height.

 A numerical simulation of a uniform flow in an open channel under wind stress conditions was performed to investigate the wind effect on the surface velocity and the sub-surface velocity profile, which is of interest in river-discharge measurements. The results of the open-channel flow show that the surface velocity and the sub-surface profile are more sensitive to wind stress as the bed friction velocity is smaller and the water depth is larger. It is also found that the magnitude of the surface drift velocity in the open channel increases monotonically with the ratio of the wind stress to the bed shear stress and reaches a maximum of 3–4% of the wind speed. This fact suggests that the wind speed correction factor in the discharge measurement should be estimated considering hydraulic and wind conditions rather than a constant value. The surface drift velocity obtained by several river observations may include not only the drift current but also other unknown effects.

 Levee breaching due to overtopping flows in rivers has become increasingly frequent, particularly during torrential rains and flood events. For the purpose of risk management, it holds significant importance to gain a comprehensive understanding of the mechanisms underlying dike failure and accurately predict this process. In this study, a three-dimensional RANS model which can simultaneously simulate surface, seepage flows and sediment transport in a curvilinear coordinate system is developed. The numerical model is applied to bar formation in a meandering channel and levee breaching to verify the model's performance. The results demonstrate that the flows and bed deformation in the meandering channel and the temporal process of dike breaching is accurately reproduced in the simulation.

 The spacing is an important consideration in spur dikes design. In meandering channels, the flow field and sediment transport are affected by both the spur dikes and channel sinuosity. The paper employed a 3D model to investigate the effect of the spacing between spur dikes on bed deformation and flow in meandering channels under different sinuosity. The results showed that severe local scour occurred near both the most upstream spur dike and spur dikes downstream of the meander apex. As the spacing increased, complete scour holes emerged starting from the downstream spur dikes, the potential bank erosion became larger, and the flow near spur tips became more turbulent with a large velocity gradient. The maximum local scour depth increased with the spacing and decreased with the channel sinuosity in most cases. The non-uniform spacing could be used to lower the construction cost, and different sinuosity required different spacing arrangements. Based on the simulation results, the performance of different spacing was quantitatively evaluated simply by considering four factors of bank protection, structural stability, construction cost, and aquatic habitat.

 Fluvial floods have been considered a lower risk to society in Taiwan since the implementation of well-protection works. However, given the impact of climate change on water-related disasters, it is necessary to re-evaluate the capability of existing facilities in Taiwan to adapt to a changing climate. By utilizing the d4PDF dataset in the rainfall-runoff simulation, we estimated potential changes in peak discharge in the Choshui River basin in central Taiwan to identify the impact of future extremes on runoff characteristics. Comparing the results with the design discharge of downstream Choshui River (100-year return period), the d4PDF +4K ensemble experiments indicate a 9.9%-27.9% increase in peak discharge, highlighting a heightened risk of fluvial flooding. Impact assessments of overflow were conducted by a 1-D river flow and a 2-D inundation model to simulate the fluvial flood inundation. To solve the intrinsic limitation of the 1-D river channel routing with mesh size for large-scale rivers, this research proposes a new scheme to automatically build up the connection between the river, floodplain, and inland cells to consist of the simulation with the actual channel boundary. This procedure successfully progresses the model simulation to be flexible to fit different shapes of river channels. Applied to the model, we provide several case studies of fluvial inundation by incorporating discharge derived from d4PDF simulation input. The results of the inundation simulation show that coastal areas around the Choshui River basin expose to a high risk of fluvial floods under climate change, highlighting the insufficiency of existing flood prevention works. Also, we demonstrate the potential risk areas for future fluvial floods, alone the Choshui River.