Future changes in sea-level pressure (SLP) around Japan are investigated using the Coupled Model Intercomparison Project phase 6 (CMIP6) projections, and their impacts on future regional change of surface air temperature and precipitation, including their extremes, are estimated based on observed statistics. The SLP patterns around Japan are quantified by defining a surface wind index, and are compared with the CMIP5 projections, revealing a similar mean change but a reduction in inter-model uncertainty. The future seasonal cycle represented by the index indicates that the spring SLP pattern will appear earlier, but the autumn pattern will be delayed. The CMIP6 models projecting stronger winds from warm (wet) areas tend to simulate a warmer (wetter) future climate over Japan, which is consistent with the statistical relationships in observations. The impact of future SLP changes on regional climate is assessed using the index based on observed relationships. The results indicate that future westerly anomalies (relative to the present day) in summer will increase mean precipitation on the Sea of Japan side of eastern and western Japan, but decrease extreme precipitation on the Pacific side of eastern Japan. The southerly anomalies in winter and autumn will increase mean and extreme precipitation over western Japan.

Historically, the Japan Sea region of Honshu has experienced some of the heaviest snowfalls in the world, though snowfalls are now decreasing due to global warming. This study examined changes in streamflow seasonality and trends in monthly runoff during the past 60 years at nine river basins located across Niigata, Yamagata, Akita and Aomori Prefectures. A streamflow seasonality index (Center Time, CT) was adopted, and meteorological stations paired with gauging stations to analyze the dependence of streamflow seasonality on air temperature. Overall, there is a strong tendency for winter and early snowmelt season flows to increase (December to March), while the peak snowmelt season flows in April are decreasing. In most cases, CT shows a trend for earlier seasonal runoff, and we confirm a strong linear relationship between CT and temperature during snow cover season (December to April). The relationship between CT and temperature appears to be stronger in the south of the study region (mean R² = 0.64) than in the north (mean R² = 0.37). In addition, the regression slope (temperature sensitivity of CT) is greater in the southern region, with an average value of –6.3 days per degree Celsius compared to –4.2 days per degree Celsius in the northern region.

Empirical flood depth damage functions for detached residential buildings were developed using 340 insurance data, whose contracts were in 14 inundation areas under the three major flood events between 2018 and 2020 in Japan. The developed flood depth damage functions have two components: the probability of damage occurrence with a sigmoid function and the conditional damage ratio with a cumulative distribution function of the standard normal distribution. The developed functions accurately reproduced the observed aggregate losses. Two-fold cross-validation confirmed that the proposed method shows a 13% error to the observed loss in NRMSE, which is 3% larger than estimated by the Ministry of Land, Infrastructure, Transport and Tourism (MLIT). Focusing on the 0.45–1.0 m inundation depth, the MLIT function, which does not consider the probability of no damage, overestimates the aggregated damage loss, but our method considering the probability of no damage has good estimation accuracy. At shallow inundation depths with some probability of no damage, the accuracy of damage estimation may be improved by considering the probability of damage. The proposed method is particularly suitable for estimating damage at shallow inundation depths because it considers the probability of no damage and is extensible to existing damage functions.

Owing to the increasing severity of flood disasters in recent years caused by climate change, flood forecasting technologies are receiving increasing attention. For real-time river-stage forecasting, hybrid methods that combine the rainfall-runoff-inundation (RRI) model and machine-learning techniques have been actively researched. This study proposes a method for appropriately selecting calculated flow-rate input points and an effective deep neural network (DNN) structure for a hybrid method that employs a DNN to forecast river water levels using flow rates calculated by an RRI model and the observed river water levels. The results show that upstream flow rates, including that of each tributary, significantly contribute toward more accurate forecasts. Moreover, a hybrid-input ensemble neural network, which combines rainfall- and flow-rate-based DNN forecasts, improves the forecast accuracy by approximately 20% for a forecast lead time of up to 24 h compared with a simple DNN model. The proposed hybrid model demonstrated better forecasting accuracy than the simple DNN, even after training on a small amount of flood data, indicating its potential for application in cases with limited past observational data.

Weather radar can observe rainfall over wide areas with higher spatial and temporal resolution than other observation methods. The Thai Meteorological Department (TMD) has installed ground rainfall stations throughout Thailand to monitor rainfall but the number of these stations is insufficient. The TMD has also installed weather radar systems but their data quality is deficient due to frequent spatiotemporal errors. In this study, we evaluate the usefulness of TMD weather radar data that include spatiotemporal errors and propose a new spatiotemporal interpolation method to address the problem of missing data. A new hybrid rainfall product for runoff analysis was constructed through the temporal interpolation of radar data based on substituting ground rainfall data for missing data and the spatial interpolation of data based on inverse distance weighting. Runoff analyses showed that the hybrid rainfall product reproduced the observed discharge more accurately than ground rainfall data, demonstrating the high applicability of the proposed product.

Iron is essential for biogeochemical processes in aquatic ecosystems, but its riverine concentration can be affected by environmental conditions. This study assessed weekly fulvic acid iron (FAFe) concentration at a single sampling site in Asahi River from 2022–2023 to explore the differences in the temporal scales. The objectives of this study were to evaluate the effects of physicochemical properties of the river on the concentration of FAFe, analyze the concentration of FAFe in spring, summer, autumn and winter, and assess the relationship between FAFe concentration and land use types of the watershed. The results indicated that physicochemical parameters, such as pH and surface water temperature (SWT) seemed to influence FAFe concentration (p < 0.05). Hydrological dynamics influenced FAFe concentration and transport, revealing an increasing trend during spring (p < 0.001) and summer (p = 0.05), with non-significant trends during autumn and winter (p > 0.05). FAFe exhibited a strong positive correlation with total organic carbon (TOC) (p < 0.001). Upland fields significantly influenced FAFe concentration (p < 0.01) through runoff with abundant NO₃^– and PO₄^3– into the river. Thus, FAFe concentration in Asahi River was influenced by pH, SWT, TOC, hydrological regime, and agricultural runoff.

In contrast to unsaturated soil conditions, advective heat transport dominates in aquifers. Consequently, groundwater flow can affect the efficiency of geothermal horizontal ground heat exchangers (HGHEs) and the development of downstream thermal plumes. In contrast to deeper geothermal systems, this has not been investigated in detail for HGHEs due to their shallow installation depths. In order to address this issue, in this study, the effects of different groundwater flow conditions on a theoretical unidirectional HGHE were evaluated using the FEFLOW^® finite element simulation system. A total of 81 variants with different groundwater levels, flow velocities and flow directions were considered in the numerical simulations.

The results showed a significant increase in HGHE efficiency with increasing groundwater level (GWL) and flow velocity. In addition, significantly higher heat gains were obtained when the groundwater flow direction was orthogonal to the HGHE. The lateral spread of the thermal plumes was greatest for orthogonal flow directions, high flow velocities and groundwater levels slightly above the HGHE.

The presented FEFLOW^® approach was suitable to investigate the effects of different hydraulic subsurface conditions on a unidirectional HGHE and should be validated with real measured data in further research.

This study investigated the dynamics of suspended sediment in the Abukuma River estuary, Japan. Seasonal samples were collected using a distributed sampling technique. Additionally, surface water samples were collected during a storm event. Across all sampling periods, the predominantly fine-grained, silica and aluminum rich suspended sediment exhibited an unimodal particle size distribution. Surface water suspended sediment concentrations were strongly correlated with cross-sectional averages validating the use of surface measurements for measuring suspended sediment. Turbidity also proved to be a reliable proxy for measuring suspended sediment concentration. The power-law relationship showed a varying degree of association between discharge and sediment load (total suspended sediment, clay, silt, and sand). During a storm event, a figure-eight hysteresis loop was observed, indicating complex sediment transport dynamics. Notably, about fifty percent of the suspended sediment particles had diameters in the 4–5 μm range. While validating the practical use of surface measurements as a representative sampling point, the study highlights the limitations of power-law models for particle size rating curves, particularly for sand, and suggests the need for alternative modeling approaches. By integrating current results with continued monitoring, a better understanding of sediment dynamics and their effects on coastal landforms can be achieved.

In this study, we developed a catchment-scale rainfall-runoff model using the vertical quasi-two-dimensional surface-subsurface flow model (quasi-2D model) and applied it to the upstream catchment of the Kamo River in Japan. The modeling of slope connections was devised to enhance computational stability, and the settings of computational cell sizes were designed according to the slope angles to reduce the computational cost while maintaining accuracy. Rainfall-runoff simulations were conducted for several floods, and the model was validated by comparing it to observational data. The quasi-2D model generally reproduced the observed river discharge well; however, it tended to overestimate the peak discharge. The present model seemed to generate excessive surface flows in mountainous areas, suggesting the need to incorporate various flow pathways, such as macropores in the soil and groundwater flows, which enhance the underground drainage capacity. The simulations for an area of 138.1 km² required less than one-tenth of the actual time, partly because of the parallel computation of two independent catchments.

Thailand’s mountainous upper Yom basin experiences frequent floods due to high river discharge. However, limited rain gauges restrict hydrological applications and flood risk assessments. Weather radar offers promise for rainfall estimation using reflectivity–rainfall (Z–R) relationships, but associated uncertainties hinder their use. Therefore, this study: i) determined the Z–R relationship for radar-rainfall estimation in the upper Yom basin; and ii) assessed the uncertainty of the determined radar-rainfall. The least-square error (LSE) method and the bias-variance correction (BVC) method were used to determine the Z–R relationships and the 95% prediction intervals (PI) for radar-rainfall estimated by both methods were calculated and compared. This showed that the LSE method tended to underestimate rainfall but gave more certainty (The 95% PI is much narrower), while the BVC gave average rainfall closer to that observed but uncertainty was much higher. Even though the LSE method does not yield an accurate average radar-rainfall, the 95% PI of the radar-rainfall estimated by this method can potentially be used to determine the range of estimated actual rainfall.

Forest management is crucially important for the sustainability of water sources in watersheds because forests are important water-related ecosystem components. Thinning is one technique that affects runoff change. For this study, we examined thinning effects on long-term runoff, based on monthly runoff and rainfall data for a coniferous forest plantation in the Hitachi Ohta watershed, Japan. Paired catchment experiments were used to evaluate water yield. Monitoring was done pre- and post-thinning during 2006–2021. In 2009, we removed 50% of the trees in catchment HV by thinning management, whereas catchment HB was left as a control catchment. The mean monthly runoff and runoff coefficient of catchment HV were higher than those of catchment HB throughout the post-thinning periods. Furthermore, 105.71 mm/year and 142.97 mm/year increases were found, respectively, in yearly water yields for the first (2010–2012) and latter (2018–2021) examination periods. About 12 years after thinning, the thinning effects are expected to continue. Results show that long-term study is necessary to assess forest management in Japan.

A novel global ensemble hydrological simulation system has been developed to enhance the accuracy of terrestrial water cycle simulations. This system integrates the Today’s Earth (TE) hydrological model with atmospheric forcing from the Nonhydrostatic Icosahedral Atmospheric Model (NICAM) - the Local Ensemble Transform Kalman Filter (LETKF), Japan Aerospace Exploration Agency (JAXA) Research Analysis (NEXRA), using 128 ensemble members to account for uncertainties in land water and energy budgets and river dynamics. The system was validated against ground-based observations of snow water equivalent (SWE) and river discharge, key components in the hydrological cycle. Despite some limitations in data overlap, the results showed reasonable correlations in certain locations, with reduced errors and biases when compared to deterministic simulations. The study revealed that ensemble mean results improve overall accuracy, especially in SWE simulations where variability among members is diminished through seasonal accumulation. However, for river discharge, the variability among ensemble members affects peak flow estimates, highlighting the need for further refinement in ensemble spread and individual member analysis. Overall, the developed system, referred to as Today’s Earth - Global (TE-Global) NEXRA, demonstrates the potential for more reliable hydrological predictions through ensemble approaches, contributing to better understanding and management of water resources.