article · Water
Deep aquifers in the Souf Valley of the Algerian Desert were evaluated to assess groundwater suitability for agricultural irrigation. Physicochemical variables from 45 deep wells were analysed using multiple irrigation water quality indices and geographic information systems. The groundwater hydrochemistry is dominated by calcium, magnesium, chloride, and sulphate ions, alongside sodium chloride facies, driven by evaporation, reverse ion exchange, and rock-water interactions. Overall, the water quality indices place the groundwater into moderate to high restriction categories, indicating that it is primarily suitable for salt-tolerant crops. To improve assessment efficiency, support vector machine regression models were calibrated and validated to forecast eight irrigation water quality indices. The machine learning model demonstrated strong predictive performance, achieving validation coefficients of determination between 0.88 and 0.95. Integrating these computational models with mapping tools provides an effective framework for groundwater management in arid and semi-arid agricultural regions.
In arid agricultural zones, relying on untested deep groundwater risks soil degradation and crop failure. This research demonstrates that machine learning models can accurately predict irrigation suitability indices from basic water measurements. This capability helps water authorities and agricultural planners detect salinity risks early and select suitable crops, supporting sustainable food production in water-stressed desert environments.
This methodology provides a tested analytical framework for agricultural authorities, irrigation consultants, and water resource managers operating in arid territories. By combining machine learning regression with geographic information systems, the approach reduces the complexity of evaluating water suitability. The research is currently an applied, validated scientific workflow, meaning commercial deployment would require packaging the predictive models into accessible software or decision-support tools for field agronomists.
AI-generated from the published abstract. Always read the original work before citing.
Irrigation has made a significant contribution to supporting the population’s expanding food demands, as well as promoting economic growth in irrigated regions. The current investigation was carried out in order to estimate the quality of the groundwater for agricultural viability in the Algerian Desert using various water quality indices and geographic information systems (GIS). In addition, support vector machine regression (SVMR) was applied to forecast eight irrigation water quality indices (IWQIs), such as the irrigation water quality index (IWQI), sodium adsorption ratio (SAR), sodium percentage (Na%), soluble sodium percentage (SSP), potential salinity (PS), Kelly index (KI), permeability index (PI), potential salinity (PS), permeability index (PI), and residual sodium carbonate (RSC). Several physicochemical variables, such as temperature (T°), hydrogen ion concentration (pH), total dissolved solids (TDS), electrical conductivity (EC), K+, Na2+, Mg2+, Ca2+, Cl−, SO42−, HCO3−, CO32−, and NO3−, were measured from 45 deep groundwater wells. The hydrochemical facies of the groundwater resources were Ca–Mg–Cl/SO4 and Na–Cl−, which revealed evaporation, reverse ion exchange, and rock–water interaction processes. The IWQI, Na%, SAR, SSP, KI, PS, PI, and RSC showed mean values of 50.78, 43.07, 4.85, 41.78, 0.74, 29.60, 45.65, and −20.44, respectively. For instance, the IWQI for the obtained results indicated that the groundwater samples were categorized into high restriction to moderate restriction for irrigation purposes, which can only be used for plants that are highly salt tolerant. The SVMR model produced robust estimates for eight IWQIs in calibration (Cal.), with R2 values varying between 0.90 and 0.97. Furthermore, in validation (Val.), R2 values between 0.88 and 0.95 were achieved using the SVMR model, which produced reliable estimates for eight IWQIs. These findings support the feasibility of using IWQIs and SVMR models for the evaluation and management of the groundwater of complex terminal aquifers for irrigation. Finally, the combination of IWQIs, SVMR, and GIS was effective and an applicable technique for interpreting and forecasting the irrigation water quality used in both arid and semi-arid regions.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.3390/w15010182
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
