ABSTRACT Reliable climate projections are essential for climate-risk assessment and planning. Yet, uncertainty in Global Climate Models (GCMs) and ensemble construction remains a major challenge. This study developed a machine learning-based selective multimodel ensemble framework to project precipitation, maximum temperature (Tmax), and minimum temperature (Tmin) over the Kangsabati, Keliaghai, Silabati, and Dwarkeswar (KKSD) river basin in eastern India. India Meteorological Department gridded observations were combined with statistically downscaled and bias-corrected NEX-GDDP-CMIP6 simulations. Multiple machine-learning algorithms were evaluated, and Extra Trees was adopted because of its strong predictive performance. A parsimonious four-GCM subset was evaluated against conventional multimodel approaches. Future projections for 2025–2098 under different SSP indicate an overall decline in annual precipitation, ranging from approximately −2.45% to −9.16% relative to 1991–2014, along with strong seasonal redistribution characterised by increased winter precipitation and persistent reductions in monsoon precipitation. Annual Tmax increases from approximately 0.39–0.44 °C in 2025–2039 to 1.53, 2.16, and 2.98 °C in 2085–2098 under SSP2-4.5, SSP3-7.0, and SSP5-8.5, respectively, while corresponding Tmin increases reach 1.40, 1.85, and 2.32 °C. The proposed framework therefore provides a parsimonious, explicitly evaluated approach to regional climate projection, while the projected climatic changes may have important implications for future water-resources planning in the KKSD basin.
ABSTRACT Four-panel summary diagram showing: (top left) maps of national adaptation plan locations in the Seomjin River Basin, (top right) boxplots of changes in reliability, resilience, and vulnerability across four water-use sectors, (bottom left) choropleth maps of internal dependency under baseline and difference scenarios for each sector, and (bottom right) scatter plots of RRV metrics against internal dependency with LOWESS trend lines. Climate-induced hydrological variability poses increasing challenges to basin-scale water security, highlighting the need to evaluate the effectiveness of climate adaptation strategies under future uncertainty. This study assesses how national adaptation plans influence water security in the Seomjin River Basin using an integrated modelling framework that couples the Soil and Water Assessment Tool and the Water Evaluation and Planning system. Water security is evaluated through system performance indicators, including reliability, resilience, and vulnerability, together with an internal water dependency metric based on the water footprint concept. The results show that adaptation measures improve recovery capacity and reduce localized supply stress, particularly for domestic and industrial water uses. In contrast, agricultural water supply and instream flow remain largely governed by hydrological variability and show limited response to adaptation. The analysis of internal water dependency indicates that the adaptation plans do not substantially alter the basin's structural reliance on internally generated water resources. These findings suggest that current adaptation strategies enhance operational performance but are insufficient to fundamentally strengthen long-term water security under climate uncertainty.
ABSTRACT Flow diagram showing rainfall-driven flood modelling using HEC-RAS, comparing baseline, nature-based solutions, structural measures, and hybrid scenarios, with reductions in flood depth, velocity, and discharge across the catchment. Rapid urban expansion and informal settlements are increasing exposure to flood hazards in urbanising catchments such as the Nairobi River Catchment. The 2024 El Niño event caused widespread flooding, affecting critical infrastructure and highlighting the urgent need for effective mitigation strategies. This study evaluates three flood management approaches using the HEC-RAS model: (1) modified channels as grey infrastructure, (2) soil organic matter enhancement and reforestation as Nature-based Solutions, and (3) a hybrid strategy. Model performance showed strong agreement with observed flows (R2 = 0.88; NSE = 0.75). Spatial datasets, including Global Forest Change and Land Productivity Dynamics, were used to identify intervention zones and parameterise scenarios. Results show that NbS reduced peak discharge by 37–75% and delayed peak flow by about one hour. Grey infrastructure achieved the greatest impact, reducing flood depths by 55–60% and peak discharge by 48–75%, but increased flow velocities. The hybrid strategy provided the most balanced hydraulic performance, combining runoff attenuation with improved channel conveyance to deliver consistent reductions in flood magnitude and enhanced flow regulation. These findings demonstrate that integrated approaches combining engineered and nature-based interventions offer a resilient pathway for managing urban flood risk under increasing climatic and developmental pressures.
ABSTRACT Four linked panels with a take-home ribbon. (1) Station diagram: Mymensingh (north-central training station, daily temperature SD 4.27 degree symbol C) linked to Rajshahi (dry western, 5.03 degree symbol C) and Chittagong (coastal south-east, 3.39 degree symbol C); solid arrows show direct transfer (trained at Mymensingh, tested at Rajshahi and Chittagong) and dashed arrows show reverse transfer (trained at Rajshahi or Chittagong, tested at Mymensingh); a small bar chart ranks native variability (Rajshahi greater than Mymensingh greater than Chittagong). (2) Six ML and five DL models over four horizons, with the best model at Mymensingh per horizon: CNN_LSTM at 1 day (RMSE 1.15 degree symbol C, R2 0.93) and 7 days (1.60 degree symbol C, 0.87), CNN_GRU at 15 and 30 days (1.63 degree symbol C, 0.86). (3) Grouped bar chart of mean R2 change across eleven models for the four transfer directions; Mymensingh-to-Chittagong shows the largest loss (up to 28% at 30 days) and Rajshahi-to-Mymensingh the smallest (under 3%). (4) Key findings: RMSE rose 6.1 to 68.2% under transfer to Rajshahi; at Chittagong RMSE fell 3.1% at 1 day through a scale effect while R2 dropped 21.5-28.1% at 7-30 days; Rajshahi-to-Mymensingh RMSE changed only -1.4 to +6.8%, with RFR leading at 7-30 days; Chittagong-to-Mymensingh RMSE rose 11.4-19.6%; XGB collapsed at 30 days at every station (RMSE 2.42 degree symbol C at Mymensingh). Ribbon: transfer difficulty follows native temperature variability, not training-station identity, and transferability should be judged by R2 rather than RMSE alone. Accurate air temperature forecasting supports agricultural planning in climate-vulnerable regions, yet spatial transferability of machine learning and deep learning models remains underexplored. This study compares eleven architectures (MLR, RFR, GBR, XGB, LGBMR, SVR, LSTM, CNN, GRU, CNN_LSTM, CNN_GRU) across four horizons (1, 7, 15, 30 days) using 25-year records (2000–2024) from three climatically distinct Bangladesh stations. Models trained at Mymensingh were validated without retraining at Rajshahi and Chittagong, then tested in reverse, training independently at each station and validating at Mymensingh. CNN_LSTM led at the 1- and 7-day horizons (RMSE = 1.15 ± 0.05 °C and 1.60 ± 0.04 °C) and CNN_GRU at 15 and 30 days (RMSE = 1.63 ± 0.05 °C and 1.63 ± 0.06 °C), all with R2 above 0.85. Transfer produced 6–68% RMSE degradation at Rajshahi, and a smaller RMSE change but larger R2 loss at Chittagong, reaching 50.7% for MLR at 30 days. Reverse transfer confirmed that difficulty tracks native variability rather than training-station identity, with milder changes and RFR outperforming the deep hybrid models at Rajshahi. XGB, though efficient, collapsed at the 30-day horizon at all stations. A seven-metric, PCA-based composite ranking confirmed these patterns, guiding horizon- and location-specific model selection, supporting climate-resilient agricultural planning in data-sparse regions.
ABSTRACT Graphical abstract showing a framework for evaluating the environmental, economic, and social impacts of inter-basin water transfers in China under future climate and socioeconomic scenarios. alt text of Graphical Abstract: Graphical abstract showing a framework for evaluating the environmental, economic, and social impacts of inter-basin water transfers in China under future climate and socioeconomic scenarios. Inter-basin water transfer (IBWT) projects have been implemented across the globe to mitigate water shortages by reallocating water resources across time and space. The environmental, economic, and social impacts arising from the inherent complexities of IBWT projects remain insufficiently explored under the effects of climate change. This study assessed the positive and negative impacts of China's IBWT network for the 2025–2099 period under multiple future climate scenarios. The results revealed that IBWT is projected to enhance the environmental flow conditions in 21 of the 38 recipient sub-basins under low-emission scenarios. Also, IBWT will potentially generate substantial economic opportunity benefits and provide the water supply for nearly 500 million people annually. However, under challenging scenarios with low water availability, some source sub-basins could face low environmental flow conditions and high socio-economic costs following water transfer. IBWT impacts are expected to be spatially heterogeneous, with the North China Plain gaining the greater benefits. Moreover, IBWT in China may introduce distributional inequalities of environmental, economic, and social impacts between source and recipient sub-basins. Our results highlight the multidimensional impacts of IBWT across future climate scenarios, underlining the need for continued adjustment of national water diversion guidelines and climate change mitigation in China.
