Water resources modeling in Wadi Numan, Western Saudi Arabia using remote sensing and GIS-based multi-criteria

Abstract

Exploring water resources is crucial for sustaining life in arid-hyperarid regions. Multicriteria derived from remote sensing, geologic, and climatic were integrated into a GIS-based data-driven frequency ratio (FR), and evidential belief function (EBF) techniques for demarcating water resources in Wadi Numan and its surroundings, west of Saudi Arabia. Remote sensing data including SRTM, Sentinel-1, and Landsat-8 have allowed for characterizing the geologic, hydrologic, structural features, and meteoric conditions of the study area. Fourteen GIS- layers include topography, slope, curvature, depression, TRI, drainage density, TWI, Distance to River, SPI, InSAR CCD, Lithology, Vegetation, lineament density, and rainfall intensity. These layers were prepared, normalized, and computed using FR and EBF, then fused and overlayed using GIS methods. The two outputs of FR and EBF revealed that the excellent areas that hold water were covering 3.63%, and 6.08%, respectively. Assessing and validation of the two models using groundwater wells and area under the curve (AUC). The AUC for validated, and all samples of FR are 0.80, and 0.82%, but EBF is 0.77, and 0.79, respectively. This has proven an accepted implementation than the EBF method. Using the InSAR CCD technique derived from Sentinel-1, two different date images revealed that areas of incoherence have high potential. Overall, implementing data-driven techniques is crucial for modeling water resources in arid regions and the applied methods are important and can be used in other areas of environmental conditions.