Spatial mapping of groundwater potential using entropy weighted linear aggregate novel approach and GIS

• Main Authors: Al Abadi, A. M., Pourghasemi, H. R., Shahid, S., Ghalib, H. B.
• Format: Article
• Published: Springer Verlag 2017
• Subjects: TA Engineering (General). Civil engineering (General)
• Online Access: http://eprints.utm.my/id/eprint/75364/; https://dx.doi.org/10.1007/s13369-016-2374-1

A novel approach for demarcation of groundwater potential by integrating entropy information theory and linear weighted aggregation method has been proposed in this study. Altun Kupri Basin in northern Iraq is selected to explain the benefits of the proposed method. Ten groundwater conditioning factors are chosen depending upon literature reviews, local conditions, and data obtainability. These layers are ground surface elevation, slope angle, plan curvature, topographic wetness index, stream power index, aquifer hydraulic head, aquifer hydraulic conductivity, depth to groundwater, total dissolved solid, and typical soil infiltration rate. Assessment of the importance of the factors in defining groundwater potential using entropy theory reveals that aquifer characteristics such as hydraulic head and hydraulic conductivity are important along with topographic factors in delineating groundwater potentiality. The hydraulic head received the highest entropy weight of 0.23, followed by hydraulic conductivity (0.17), elevation (0.14), and topographic wetness index (0.13). The other factors received entropy weights less than 0.1, which mean that they less contribute in groundwater availability in the basin. The groundwater potential index was constructed using a linear weighted aggregation of all the ten factors. Values of the groundwater potential index are classified into five classes using Jenks classification scheme: very low, low, moderate, high, and very high. Map of potential index was validated using relative operating characteristic curve. The area under relative operating characteristic is found to be 0.74, and therefore, the prediction accuracy was good.