Groundwater contamination with sub-lethal dissolved contaminants poses significant health risks globally, especially in rural India, where access to safe drinking water remains a critical challenge. This study explores the hydrogeochemical characterization and associated health risks of groundwater from shallow aquifers in the Marginal Ganga Alluvial Plain (MGAP) of northern India. The groundwater chemistry is dominated by Ca-Mg-CO₃ and Ca-Mg-Cl types, where there is dominance of silicate weathering and the ion-exchange processes are responsible for this solute composition in the groundwater. All the ionic species are within the permissible limits of the World Health Organization, except fluoride (F⁻) and nitrate (NO₃⁻). Geochemical analysis using bivariate relationships and saturation plots attributes the occurrence of F⁻ to geogenic sources, primarily the chemical weathering of granite-granodiorite, while NO₃⁻ contaminants are linked to anthropogenic inputs, such as nitrogen-rich fertilizers, in the absence of a large-scale urban environment. Multivariate statistical analyses, including hierarchical cluster analysis and factor analysis, confirm the predominance of geogenic controls, with NO₃⁻-enriched samples derived from anthropogenic factors. The spatial distribution and probability predictions of F⁻ and NO₃⁻ were generated using a non-parametric co-kriging technique approach, aiding in the delineation of contamination hotspots. The integration of the USEPA human health risk assessment methodology with the urbanization index has revealed critical findings, identifying approximately 23% of the study area as being at high risk. This comprehensive approach, which synergizes geospatial analysis and statistical methods, proves to be highly effective in delineating priority zones for health intervention. The results highlight the pressing need for targeted mitigation measures and the implementation of sustainable groundwater management practices at regional, national, and global levels.

Modelling at the basin scale offers crucial insights for policymakers as they make decisions regarding the optimal utilization of water resources. This study employed the MIKE HYDRO Basin model to analyse water demand and supply dynamics in the Parvati Basin of Rajasthan, India, for the period 2005–2020. The MIKE11 NAM model showcased strong alignment between simulated and observed runoff during both the calibration (NSE = 0.79, PBIAS = −2%, R² = 0.79, RMSE = 4.95, RSR = 0.5, and KGE = 0.84) and validation (NSE = 0.67, PBIAS = −12.4%, R² = 0.68, RMSE = 8.3, RSR = 0.62, and KGE = 0.67) phases. The MIKE HYDRO Basin model also exhibited excellent agreement between observed and simulated reservoir water levels, with R², NSE, RMSE, PBIAS, RSR, and KGE values of 0.86, 0.81, 3.87, −2.30%, 0.43, and 0.88, respectively. The MIKE HYDRO Basin model was employed to create six distinct scenarios, considering conveyance efficiency, irrigation method, and conjunctive water use, to assess irrigation demands and deficits within the basin. In the initial simulation, featuring a conveyance efficiency of 45%, flood irrigation, and no groundwater utilization, the average water demand and deficit throughout the study period were estimated as 43.15 MCM and 3.45 MCM, respectively, resulting in a sustainability index of 0.506. Enhancing conveyance efficiency to 75% under flood irrigation and 5% conjunctive use could elevate the sustainability index to 0.92. Transitioning to sprinkler irrigation and a lift irrigation system could raise the system’s sustainability index to 1. These developed models hold promise for real-time reservoir operation and irrigation planning across diverse climatic conditions and varying cropping patterns.

Middle East (ME) countries have arid and semi-arid climates with low annual precipitation and considerable geographical and temporal variability, which contribute to their extremely erratic rainfall. The generation of timely and accurate climatic information for the ME is anticipated to be aided by global reanalysis products and satellite-based precipitation estimations. Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN) and Climate Hazards Group Infra-Red Precipitation (CHIRPS) on Google Earth Engine (GEE) were used to study rainfall in eleven chosen ME counties from 2000 to 2023. This study shows that Saudi Arabia (509.64 mm/December–January–February; DJF), Iraq (211.50 mm/September–October–November; SON), Iran (306.35 mm/SON), Jordan (161.28 mm/DJF), Kuwait (44.66 mm), Syria (246.51 mm/DJF), UAE–Qatar–Bahrain (28.62 mm/SON), Oman (64.90 mm/June–July–August; JJA), and Yemen (240.27 mm/SON) were the countries with the highest rainfall. Due to improved ground station integration, CHIRPS also reports larger rainfall anomalies, with a peak of 59.15 mm in DJF, mainly in northern Iran, Iraq, and Syria. PERSIANN understates heavy rainfall, probably because it relies on infrared satellite data, with a maximum anomaly of 4.15 mm. Saudi Arabia saw heavy rain during the JJA months, while others received less. More accurate rainfall forecasts in the ME can lessen the effects of floods and droughts, promoting environmental resilience and regional economic stability. Therefore, a more comprehensive understanding of all the relevant components is necessary to address these difficulties. Both environmental and human impacts must be taken into account for sustainable solutions.

Artificial neural networks (ANNs) offer considerable advantages in predicting evaporation (EVAP), particularly in handling nonlinear relationships and complex interactions among factors like soil surface temperature (SST) and wind speed (WS). In Al Medina, Saudi Arabia, the connections among WS, SST at 5 cm, SST at 10 cm, and EVAP have been modeled using an ANN. This study demonstrates the practical effectiveness and applicability of the approach in simulating complex nonlinear dynamics in real-life systems. The modeling process employs time series data for WS, SST at both 5 cm and 10 cm, and EVAP, gathered from January to December (2002–2010). Four ANNs labeled T1–T4 were developed and trained with the feedforward backpropagation (FFBP) algorithm using MATLAB routines, each featuring a distinct configuration. The networks were further refined through the enumeration technique, ultimately selecting the most efficient network for forecasting EVAP values. The results from the ANN model are compared with the actual measured EVAP values. The mean square error (MSE) values for the optimal network topology are 0.00343, 0.00394, 0.00309, and 0.00306 for T1, T2, T3, and T4, respectively.