The Genale Dawa River Basin (GDRB) in southeastern Ethiopia is highly vulnerable to climate variability. However, limited research has examined how well the latest CMIP6 climate models simulate key climate variables in this complex, data-scarce region. This study evaluates 12 CMIP6 Global Climate Models (GCMs) in reproducing historical precipitation, maximum temperature (Tmax), and minimum temperature (Tmin) over the GDRB for 1985-2014, and projects future changes for 2021-2050 and 2051-2080 under SSP2-4.5 and SSP5-8.5 scenarios. Observational data from ENACTS and 60 ground-based stations were used to validate model outputs, which were resampled to 0.05° × 0.05° resolution and bias-corrected using quantile mapping. Model performance was evaluated using Mean Bias Error (MBE), Root Mean Square Error (RMSE), Pearson correlation, and Taylor Skill Score (TSS). The multi-model ensemble (MME) outperformed individual GCMs, achieving TSS values of 0.80 for precipitation, 0.98 for Tmax, and 0.99 for Tmin. Among individual models, CNRM-CM6-1 performed best for precipitation, GFDL-CM4 and NorESM2-MM for Tmax, and FGOALS-g3 and MRI-ESM2-0 for Tmin. Most models reproduced observed spatial and temporal patterns well but showed a systematic cold bias in Tmin, while Tmax was more accurately simulated. Under the SSP5-8.5 mid-century scenario, Tmax and Tmin are projected to increase by about 1.8°C, and precipitation by 11.1% (5.4mm). These findings demonstrate the value of ensemble-based projections and robust model evaluation for improving climate risk assessment and adaptation planning in the GDRB.

Hybrid GR4J-LSTM modeling for streamflow prediction of extreme events in data-scarce regions: Upper Blue Nile Basin, Ethiopia

A scalable framework for soil property mapping tested across a highly diverse tropical data-scarce region

Rainfall disaggregation is a key challenge in hydrology, especially in regions with limited high-resolution records. This study applies the Random Bartlett–Lewis Rectangular Pulse Model to four regions of Hellas to generate hourly rainfall from daily totals. The work is novel in evaluating the model under data-scarce Mediterranean conditions, incorporating a two-tiered uncertainty analysis, testing alternative pulse intensity distributions (Gamma and Exponential), and comparing its performance with a deterministic machine learning (ML) approach. Results show that the RBLRPM reproduces essential rainfall properties such as variance, autocorrelation, skewness, and dry spell probabilities, even when calibrated with as little as three years of data. The ML approach ensures perfect conservation of daily totals and computational efficiency, but it smooths temporal variability and underestimates extremes. By contrast, the stochastic RBLRPM captures clustering, intermittency, and heavy tails more realistically, which is crucial for hydrological design and flood risk analysis. The Gamma distribution consistently outperforms the Exponential form, though both remain applicable. Overall, the Gamma-based RBLRPM offers a robust and transferable method for rainfall disaggregation in data-limited contexts, highlighting the importance of stochastic approaches for water resource management, infrastructure resilience, and climate adaptation.

Mapping of rice-cropping regimes is crucial for effective irrigation planning and yield monitoring, particularly in data-scarce regions. We analyzed 48 months of 3 m PlanetScope NDVI data, aggregated to a 25 m hexagonal grid, and used Dynamic Time Warping Clustering to segment phenological patterns. Internal validation consistently identified two main clusters, indicating two dominant seasonality modes. Cluster 1 exhibited a higher mean NDVI, fewer low-canopy months, more vigorous growth periods, more peaks, and greater annual cycling, which suggests irrigated double cropping. Cluster 2 exhibited prolonged low NDVI values and a greater amplitude, consistent with single-rainfed systems. The rain–NDVI analysis supported these findings: Cluster 1 responded modestly to rainfall, whereas Cluster 2 exhibited a stronger and delayed response. Independent spatial checks confirmed these classifications. Off-season greenness, measured as NDVI above 0.50 from July to November, was concentrated near main and secondary canals and decreased with distance from intake points. This workflow combines DTW clustering with rainfall lag and off-season greenness analysis, effectively distinguishing between irrigated and rain-fed regimes using satellite time series. These findings are considered indicative rather than definitive, providing an assessment of cropping systems in Timor-Leste and demonstrating that DTW-based NDVI clustering offers a scalable approach in data-scarce regions.

Accurate climate data is essential for planning climate resilience and understanding regional climate variability, particularly in high-altitude, data-scarce regions such as Nuwara Eliya, Sri Lanka. This study evaluates the performance of NASA POWER reanalysis data (0.5° latitude × 0.625° longitude resolution) by comparing it with ground-based meteorological observations from 2009 to 2021, addressing a gap that remains largely unexplored in the existing literature. Key climate variables, including precipitation, temperature, wind speed, and relative humidity (daily data), along with cloud cover (monthly data, 2017–2020), were validated using the correlation coefficient (R), mean error (ME), root mean square error (RMSE), and percent bias (PBIAS). The results showed strong agreement for temperature (bias-corrected R = 0.66), wind speed (R = 0.72), cloud cover (R = 0.92), and relative humidity (R = 0.63). Precipitation analysis revealed that dry-day overestimation drove positive biases in FIM, SIM, and NEM, whereas SWM biases resulted from underestimation of monsoon rainfall intensity. The Mann-Kendall test and Sen’s slope estimator were applied to detect long-term trends (1990–2020), revealing a significant increase in minimum temperature (0.01 °C/year) and shortwave radiation (0.2 W/m²/year), and a decline in wind speed (−0.03 m/s/year), indicating significant climate shifts in the region. This study provides a comprehensive validation of NASA POWER data in Sri Lanka’s central highlands, underscores the need for localised bias correction, and demonstrates its potential for broader application in similar high-elevation, data-constrained regions for climate impact assessment and resource planning.

Droughts severely impact agriculture, food security, and water resources, particularly in data scarce regions like the Genale Dawa River Basin (GDRB). This study evaluates the performance of five satellite rainfall products including CHIRPSv2.0, RFE2.0, TAMSATv3.1, PERSIANN-CDR, and ARC2 over 2001-2020 using metrics such as Mean Bias, Absolute Error (MAE), Root Mean Square Error (RMSE), Correlation Coefficient (CC), Probability of Detection (POD), False Alarm Ratio (FAR), and Critical Success Index (CSI). Furthermore, the studies examined the relationship between climate indices including the Pacific Decadal Oscillation (PDO), North Atlantic Oscillation (NAO), El Niño Southern Oscillation (ENSO), and Atlantic Multidecadal Oscillation (AMO) and drought variability. CHIRPSv2.0 demonstrated the highest accuracy, especially at Ginir (CC = 0.98) and Chewbet (CC = 0.75), and the lowest errors at Teferekella (MAE = 45, RMSE = 65). Based on a multi-criterion ranking approach, CHIRPSv2.0 was identified as the most reliable satellite-based product for daily rainfall estimation across the entire basin, exhibiting strong performance metrics (Lp = 0.433, CC = 0.84, POD = 0.85, CSI = 0.46). CHIRPSv2.0 was used to analyze drought spatiotemporal patterns using SPI-3 and SPI-6. Drought severity, frequency, and duration were highest in the west-central and northern GDRB, with major events occurring in 1988, 1991-1993, 2000, 2004, 2011, and 2012. Climate index analysis showed that AMO and NAO positive phases were associated with wetter conditions, while negative PDO and ENSO phases corresponded with drier periods, especially in central and eastern areas. These findings highlight CHIRPSv2.0's reliability for drought monitoring and its value for early warning and mitigation planning in the GDRB.

High-quality meteorological data are essential for climate monitoring and renewable energy applications. ERA5-Land, a newly released high-resolution reanalysis dataset, provides a wide range of meteorological variables, but its accuracy remains a concern. This study evaluated the performance of ERA5-Land in the Lower Jinsha River Basin, the largest clean energy base in China, focusing on precipitation, wind speed, air temperature, and solar radiation. A statistical bias correction procedure was developed, combining month-specific regression fitting with daily and hourly adjustments. Results indicated that air temperature estimates agreed best with ground observations, with a coefficient of determination (R2) exceeding 0.87 and percent bias (Pbias) below 15%, followed by solar radiation. Precipitation and wind speed, in contrast, exhibited larger uncertainties (R2 < 0.31, Pbias up to 67.76%). After applying the statistical bias correction, systematic biases were largely eliminated across all examined variables. Absolute errors decreased by more than 10%, and temporal consistency also improved moderately, especially for wind speed and solar radiation, where R2 increased by 29.5% and 25.8%, respectively. The corrected dataset captured basin-wide climatic variations from 1980 to 2019, including decreasing precipitation, increasing temperature and solar radiation, and the spatial heterogeneity changes in wind speed. Overall, this study contributes to better knowledge of ERA5-Land uncertainties in multiple meteorological variables and provides a practical statistical correction framework, which can serve as a reference for data-scarce regions with similar climatic and geographical conditions and clean energy development contexts.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-025-24942-4.

Accurate forecasting in heterogeneous spatiotemporal environments requires models that are both generalizable and interpretable, while also preserving cross-scale symmetry between temporal and spatial patterns. Existing deep learning approaches often struggle with limited adaptability to data-scarce regions and lack transparency in capturing cross-scale causal factors. To address these challenges, we propose a novel framework, Cross-Scale Symmetry-Aware Causal Spatiotemporal Modeling with Adaptive Fusion and Region-Knowledge Transfer, which integrates three key innovations. First, a Dynamic Spatio-Temporal Fusion Framework (DSTFF) leverages frequency-aware temporal transformations and adaptive graph attention to capture complex multi-scale dependencies, ensuring temporal–spatial symmetry in representation learning. Second, a Region-Knowledge Enhanced Transfer Learning (RKETL) mechanism distills knowledge across regions through teacher–student distillation, graph-based embeddings, and meta-learning initialization, thereby maintaining structural symmetry between data-rich and data-scarce regions. Third, a Multi-Granularity Causal Inference Prediction Module (MCIPM) uncovers cross-scale causal structures and supports counterfactual reasoning, providing causal symmetry across daily, weekly, and monthly horizons. Comprehensive experiments on multi-regional logistics datasets from China and the U.S. validate the effectiveness of our approach. Across six diverse Chinese regions, our method consistently outperforms state-of-the-art baselines (e.g., PatchTST, TimesNet, FEDformer), reducing MAE by 18.5% to 27.4%. On the U.S. Freight dataset, our model achieves significant performance gains with stable long-horizon accuracy, confirming its strong cross-domain generalization. Few-shot experiments further demonstrate that with only 5% of training data, our framework surpasses the best baseline trained with 20% data. Robustness analyses under input perturbations and uncertainty quantification show that the model maintains low error variance and produces well-calibrated prediction intervals. Furthermore, interpretability is concretely realized through MCIPM, which visualizes the learned causal graphs and quantifies each regional factor’s contribution to forecasting outcomes. This causal interpretability enables transparent understanding of how temporal spatial dynamics interact across scales, supporting actionable decision-making in logistics management and policy planning. Overall, this work contributes a unified spatiotemporal learning framework that leverages symmetry principles across scales and regions to enhance interpretability, transferability, and forecasting accuracy.

Accurate estimation of reference evapotranspiration (ET0) is crucial for irrigation planning and sustainable water management. This study developed Artificial Neural Network (ANN) models to simulate daily FAO-56 Penman-Monteith ET0 using limited meteorological inputs at Junagadh station, Gujarat, India. The Gamma Test was employed to identify optimal input combinations, revealing that maximum temperature (Tmax), wind speed (WS), solar radiation (SR), and relative humidity (RHmean) were the most influential predictors. ANN models with various input configurations (one to six variables) were trained and evaluated using statistical indicators such as RMSE, R², NSE, MAPE, and Willmott’s Index (WI). Results showed that the ANN model with three inputs (Tmax, WS, SR) achieved RMSE = 0.4722 mm/day, R² = 0.9463, NSE = 0.9029, and MAPE = 9.70%, while the four-input model (Tmax, RHmean, WS, SR) yielded RMSE = 0.4504 mm/day, R² = 0.9652, NSE = 0.9116, and MAPE = 8.56%. Models with more than four inputs offered only marginal improvement, indicating that three or four parameter combinations provide optimal accuracy with computational efficiency. The findings confirm that ANN can reliably replicate the nonlinear dynamics of ET0 and serve as a viable alternative to the FAO-56 PM method in data-scarce regions, offering accurate and efficient ET0 prediction for irrigation scheduling and water resource planning. The developed ANN model can serve as an efficient decision-support tool for irrigation scheduling and water resource management in arid and semi-arid regions with limited climatic data availability.

The daily mean temperature prediction is essential to implement agricultural adaptation and disaster risk reduction strategies in countries with varied climatic regions like Pakistan. Traditional machine learning methods often had difficulty complying with thermodynamic constraints, limiting their practicality for temperature estimation. The study proposed a Physics-Informed Neural Network (PINN), which integrates data and governing thermal principles to overcome these limitations. Integrating thermal equilibrium conditions within the loss function inherently consolidates the thermodynamic coherence of the construction. The high-resolution meteorological data from the Badin, Dadu and Rohri observation networks are used to build domain-specific features. These include diurnal excursions of temperature and humidity, a cyclic year encoding and wind-humidity ratios that capture the nonlinear mesoscale thermodynamics of the system. The PINN shows strong predictive ability as compared to a benchmark linear regression, some ensemble algorithms, and feedforward networks (R2 = 0.9975 Badin, 0.9974 Dadu, 0.9949 Rohri). SHAP and LIME, used in feature importance quantification, help to identify temperature drivers. In Badin, wind regimes have the most influence, whereas in Dadu and Rohri, lingering time trends have the most impact. With a focus on physical plausibility and explainable AI, the proposed methodology combines the probabilistic advantages of statistical learning with the constraint-based approach of atmospheric physics. This leads to resilient and spatially flexible predictions of temperature in data-scarce regions. As the study shows, PINNs could become a game-changing operational meteorological forecast technology when observational networks are weak or lacking altogether.

This study integrates geospatial modeling with multi-criteria decision analysis for an improved approach to landslide susceptibility mapping (LSM). This approach addresses key challenges in LSM through sophisticated multicollinearity analysis and machine learning strategies. We compared three machine learning models for weighting, and of them the Permutation-Weighted model yielded the best prediction results, with an Area Under Curve (AUC) of 95%, an accuracy of 69%, and a recall of 66%. To resolve perfect multicollinearity (r = 1) between land use land cover (LULC) and geological factors, we implemented Principal Component Analysis (PCA). The selected factors demonstrated strong predictive power, with the PCA-derived features exhibiting the best performance, having a Variation Inflation Factor (VIF) of 1.004. Slope appeared as the most influential factor (51.7% contribution), while the Topographic Wetness Index (TWI) was less dominant with only 6.6%. Multiple landslide susceptibility mapping methods yielded consistent results, with 29.8–30.1% of the study area showing moderate susceptibility and 35.2–36.9% in the high to very high susceptibility class. The model also incorporated vulnerability parameters weighted by the United Nations Office for Disaster Risk Reduction (UNDRR) indicators, including farmland, buildings, bare land, water bodies, roads, and amenities to generate hazard, vulnerability, and risk maps. The results were verified through visual comparison with high-resolution Google Earth imagery. The Permutation-Weighted model performed better than others, categorizing 12.4% at high-risk, while Random Forest (RF) categorized 7.2% at high risk. This study makes three key contributions: (1) It establishes the effectiveness of PCA/VIF for variable selection, (2) it provides a comparison of machine learning weighting techniques, and (3) it validates a workflow applicable to data-scarce regions.

Agricultural sustainability in flood-prone and physiographically diverse regions of Northeastern India is constrained by limited knowledge of land and soil suitability. This study addresses a critical research gap by integrating geospatial techniques with soil analysis to assess soil suitability and crop productivity through detailed mapping of physiographic units and associated soil characteristics in the Jiyadhol watershed, Northeast India. Four major landforms, piedmont plain, upper alluvial plain, lower alluvial plain, and flood plain, were delineated using remote sensing data, and 170 surface soil samples were analyzed for key physicochemical properties. Indices, such as the soil productivity index (12.1-61.7), potential productivity index (25.7-77.2), and crop yield index (0.41-2.76), were derived to assess suitability for major rice ecosystems: Sali rice (Aman), Ahu rice (Aus), and Boro (summer rice). The normalized difference vegetation index (NDVI) during the vegetative stage showed a strong positive correlation with organic matter and crop yield. Piedmont plain soils were found to be unsuitable for rice due to poor fertility and erosion, but they hold promise for oilseeds and pulses under improved management. The results highlight an inverse relationship between soil loss rates and productivity indicators, indicating the urgency of erosion control. Moreover, an integrated GIS and field data-based soil suitability map for rice generated in this study serves as a vital decision-support tool for farmers, planners, and policymakers. Overall,this research advances sustainable and climate-resilient agricultural planning in flood-affected, data-scarce regions like Northeastern Indiaby providing a spatially explicit framework for assessing agro-potential.

Reliable precipitation inputs are essential for hydrological modeling, yet global precipitation products often exhibit substantial discrepancies that introduce significant uncertainties into streamflow simulations and projections. In this study, we assessed the relative contribution of precipitation dataset uncertainty to discharge simulations and projections, in comparison with uncertainties from model structure, model parameters, and climate projections, in the Liujiang catchment, southwest China. Three widely used satellite-based products (CHIRPS, PERSIANN, and IMERG) and one reanalysis dataset (ERA5) were combined with three hydrological models of varying structural complexity to simulate streamflow. Using an ANOVA-based variance decomposition framework, we quantified the contributions of different uncertainty sources under both historical and future climate conditions. Results showed that precipitation input uncertainty dominates discharge simulations during the calibration period, contributing over 60% of total variance particularly at high flows, while interactions among precipitation, model structure, and parameters govern low-flow simulations. Under future climate scenarios, climate projection uncertainty overwhelmingly dominates discharge predictions with 50–80% of uncertainty contribution, yet precipitation products still contribute significantly across time scales. The compensation of precipitation biases by hydrological models can cause parameter values to deviate from their true physical meaning. This deviation may further amplify the differences in discharge projections driven by different precipitation products under future climate conditions and increase the overall uncertainty of streamflow projections. Overall, this study introduced an integrated approach to simultaneously assess precipitation uncertainty across flow regimes and future climate scenarios. These results emphasized the necessity of using ensemble approaches that incorporate multiple precipitation products in hydrological forecasting and impact studies, particularly in data-scarce regions reliant on global datasets.

Discharge is a critical parameter in situ to calibrate and validate water quantity and quality in rivers flowing into marine environments. Hydrological modelling is a key instrument to complete the quantitative assessment of river catchment ecology and to enumerate the impacts of pollution loads into the sea. Nutrient loads from river catchments are of particular importance for enclosed marine environments such as the Black Sea. Open-source tools and instruments, promoted for use within several European-supported projects (such as FP7 enviroGRIDS, IASON, and H2020 DOORS), were applied to data-scarce regions around the Black Sea. To mitigate issues with unavailable or low-quality discharge data, determination through satellite remote sensing for the case of the Rioni River is reported in this communication as a proving ground for this novel approach. In particular, the open-source Soil and Water Assessment Tool (SWAT, ArcSWAT, Arnold et al., 1998) was employed to set up the hydrological model for the Rioni River Basin. Global 30 m resolution land cover, 30 m Global DEM, and FAO soils cover data (complemented with the national soils at 1:500,000 scale), combined with globally available climate datasets in ArcSWAT input format, were applied to set up and run the hydrological model for the Rioni Catchment. Weak quality (and lack of public availability) of hydrological discharge measurements did not allow the calibration and validation of the water quantity model. To compensate for the lack of quality in in situ discharge data, a conceptual validation was performed to apply remote sensing to address the scarcity of discharge data. Due to a satisfactory visual fit of the in situ measured and microwave satellite observation data, it was speculated that, instead of the use of a global hydrological model to derive absolute values for river discharge time series from satellite observations, one could combine microwave satellite data (available in relative values; Brakenridge et al., 2012) with absolute figures obtained via the 'at-many-stations hydraulic geometry' (AMHG) river-width-based methodology, described in the literature (Gleason and Smith, 2014), in order to recalculate relative values of satellite measurement time series into absolute values for river discharges. It was demonstrated that the high-resolution instruments of the European Sentinel satellite series would allow for the measurement of the required parameters for wide rivers such as the Rioni, thus sensing discharge data remotely, supporting calibration and validation of the hydrological model set-up. This may contribute to resolving environmental governance challenges of river discharge and nutrient loads data availability in the Black Sea Basin. Keywords: discharge remote sensing, river basin, catchment hydrology, Black Sea.

Land use land cover (LULC) mapping using deep learning greatly enhances our understanding of geographic, socioeconomic, and urban development patterns. However, annotated satellite data are scarce in developing countries in South and East Asia due to funding and urban heterogeneity. We present the Bangladesh Open LULC Map (BOLM), a high-resolution dataset with pixel-level annotations across eleven classes for the Dhaka metropolitan area and surrounding regions (4,392 hbox {km}^2, 891 million pixels). Annotations were generated using high-resolution Bing imagery following a rigorous, multi-stage validation process supported by GIS experts. Although Bing data provides fine spatial detail for visual interpretation, multispectral data remain essential for spectral analysis and temporal monitoring. We benchmarked DeepLabV3+, HRNetv2, U-net, UnimatchV2 and Segmenter ViT-16 on five merged LULC classes, achieving an overall IoU of 0.50 and F1 score of 0.66 with the UnimatchV2. Further experiments using Sentinel-2A data highlight the trade-off between spatial and spectral resolution. Our results show that while DeepLabV3+ and HRNet are effective for LULC mapping, other models such as U-Net, Segmenter, and UniMatchV2 perform even better, highlighting the model-agnostic nature of our study. The BOLM dataset can support future deep learning models and domain adaptation for LULC tasks in data-scarce regions, addressing the critical gap of high-quality LULC data in South and East Asia.

The state of Bahia in Brazil possesses significant, yet underexploited, iron ore reserves. To support the initial stages of mineral exploration in this vast region, cost-effective and rapid large-scale mapping methods are essential. This paper presents a workflow based on publicly available remote sensing data for a state mineral prospectivity mapping (MPM) for iron. The methodology employs a Random Forest (RF) classification model on Sentinel-2 multispectral images, trained with a randomly selected dataset in the image at varying distances defined from the location of known iron mines in the state. The Spectral Angle Mapper (SAM) algorithm was used to categorize the samples according to spectral similarity features with laboratory-confirmed ore signatures from samples collected in the mine pit area. The resulting MPM successfully delineated known iron districts and highlighted new, unexplored areas with potential. A quantitative evaluation of the model yielded an overall accuracy of 69.8%, a macro-average F1-score of 0.697, and a Cohen’s Kappa coefficient of 0.623, indicating a reasonable agreement beyond random chance. This work demonstrates a validated, low-cost, and simple approach for regional-scale MPM, offering a valuable reconnaissance tool for preliminary exploration, particularly in extensive and data-scarce regions.

This study introduces the WiBISS model, a simulation tool designed to assess the economic and epidemiological impact of a hypothetical African Swine Fever (ASF) vaccination in wild boar in Northern Italy. Using real ASF outbreak data from January 2022 to June 2024, the model evaluates how prompt vaccination could reduce disease spread and economic losses. WiBISS integrates three modules: vaccination simulation, restriction zone estimation, and economic impact analysis. The first two use custom-built cellular automata (CA) in Python and ArcGIS Pro, modeling each ASF case as a cell that can be in one of three states: unvaccinated, infected, or vaccinated. Weekly iterations over 2.5 years simulate ASF progression and vaccination impact based on localized interactions and a defined vaccination radius. Three vaccination scenarios were tested: (1) a non-vaccination baseline; (2) an “ideal” scenario with immediate, 100% vaccination; and (3) multiple “realistic” scenarios with an 8-week delay and varied vaccination rates (25–75%) and radii (10–50 km). The most effective realistic scenarios (e.g., 75% vaccination rate, 50 km radius) showed a total loss of €601,800, close to the ideal scenario. WiBISS prioritizes usability over epidemiological complexity, omitting detailed virus transmission modeling to enhance applicability in data-scarce regions. Unlike detailed stochastic models, WiBISS offers rapid, economically grounded insights to guide initial outbreak response and resource allocation. Although it does not include domestic pigs due to differing transmission dynamics, WiBISS lays a foundation for phased, integrated wildlife vaccination planning that balances economic feasibility with ecological realism.

Abstract. This work presents an efficient graph-reconstruction-based approach for generating physical sewer models from incomplete information, addressing the challenge of representing the sewer drainage effect in urban pluvial flood simulations. The approach utilizes graph-based topological analysis and hydraulic design constraints to derive gravitational flow directions and nodal invert elevations in decentralized sewer networks with multiple outfalls. By incorporating linearized programming formulation to solve reconstruction problems, this approach can achieve high computational efficiency, making it suitable for application to city-scale sewer networks with thousands of nodes and links. Tested in Yinchuan, China, the approach integrates with a 1D/2D coupled hydrologic–hydrodynamic model and accurately reproduces maximum inundation depths (R2=0.95) when the complete network layout and regulated facilities are available. Simplifications, such as the adoption of road-based layouts and the omission of regulated facilities, can degrade simulation performance for extreme rainfall events compared to calibrated equifinal methods. However, design rainfall analysis demonstrates that the physical reconstruction approach can reliably outperform equifinal methods, achieving reduced variation and higher accuracy in simulating inundation areas. However, proper configuration of regulated facilities and network connectivity remains crucial, particularly for simulating local inundation during extreme rainfall. Thus, it is recommended that the proposed algorithm be integrated with targeted field investigations to further improve urban pluvial flood simulation performance in data-scarce regions.

Reliable flood monitoring and prediction remain a challenge in data-scarce regions, particularly in arid and semi-arid environments. This study explores the integration of remote sensing data and machine learning techniques to improve flood detection and early warning capabilities in Lodwar Town of the Turkwel Basin, Kenya. This depended on finding a relationship between daily rainfall and Normalized Difference Water Index (NDWI). Among multiple rainfall products evaluated, Climate Hazards Group InfraRed Precipitation with Station (CHIRPS) was selected due to its fine spatial resolution and performance. Daily NDWI time series derived from Aqua MODIS (Moderate Resolution Imaging Spectroradiometer) imagery were used as a proxy for water accumulation and flood indicators. A python-based Decision Tree Regressor (DTR) model was trained using the daily CHIRPS rainfall data with various lag times, along with auxiliary meteorological variables including relative humidity, wind speed, and mean temperature for the period from 2002 to 2024 to predict NDWI of Lodwar Town. The machine learning model substantially improved the correlation between rainfall and NDWI, raising the correlation coefficient by 25%. Spatial analysis of rainfall-NDWI correlation revealed that areas in the west, northwest, and southwest of Lodwar Town, with elevations between 508 m and 648 m have high correlation. Rainfall in these regions can serve as signal for potential rapid flooding with 0-day lag-time in Lodwar Town situated at an elevation of approximately 500 m. These areas are not necessarily the primary high rainfall sources, rather they act as signal zones for floods of Lodwar Town that can provide flood early warning information. The proposed methodology in this study can offer a practical approach to anticipatory action and flood risk reduction for vulnerable communities in remote regions with no or limited hydrometeorological stations.