A GIS-based combining of frequency ratio and index of entropy approaches for mapping groundwater availability zones at Badra–Al Al-Gharbi–Teeb areas, Iraq

Abstract

This study focus on determining the groundwater availability zones in an arid region, Iraq using bivariate frequency ratio and combining frequency ratio and index of entropy approaches linked with remote sensing and GIS techniques. For building models, an inventory of boreholes with high flow rate (8 l/s) was firstly prepared and divided into two sets training and testing along with different groundwater occurrence factors. Selection of the factors was based on availability of data, literature review, and expert opinion. The selected factors were elevation, slope angle, curvature, aspect, topographic wetness index, stream power index, geology, soil, land use/land cover, and distance to faults. The statistical relationships between groundwater occurrence factors and geographic borehole locations were investigated using likelihood ratio. The linear combination technique was first used to derive groundwater availability zones with assumption that all groundwater factors have the same influence on the groundwater availability. In the second model, the weight for each groundwater factor was calculated using index of entropy method, and thus a weighted linear combination technique was used to derive groundwater availability zones. The final groundwater availability index produced by applying both methods were classified into five classes based on Jenks classification scheme: very low, low, moderate, high, and very high. The areas covered by very low to low groundwater availability zones occupy 70 and 72 % from the total area for frequency ratio and combining frequency ratio and index of entropy models, respectively, indicating that the groundwater availability condition is low. Validation of the prospecting maps using relative operating characteristic curves indicated that the prediction rates for frequency ratio model and combining frequency ratio-index of entropy model were 0.804 and 0.806, respectively implying that the combining frequency ratio-index of entropy model was slightly better than frequency ratio model alone.