Evaluation of Frequency Ratio and Weight of Evidence Models in Spatial Mapping of Potential Groundwater Using GIS in Esfarayen county

Amirhosein Ghorbani ¹ ( MSc. Student of Survey Engineering- Geographic Information Systems, Lamei Gorgani Institute of Higher Education, Gorgan, Iran )
Mohsen Zabihi ² ( Ph.D., Department of Watershed Management Engineering, Faculty of Natural Resources, Tarbiat Modares University, Tehran, Iran )
Raoof Mostafazadeh ³ ( Associate Professor, Department of Watershed Management and Member of Water Management Research Institute, University of Mohaghegh Ardabili, Ardabil, Iran )

Submited date : 2022-12-06 Accepted date : 2022-12-20

Keywords: Geographic Information System, Groundwater exploitation, qanat, Weight of evidence, spatial modeling,

Abstract :

Optimum utilization of groundwater water resources, in a sustainable manner using Qanats, has always been of interest in arid and semi-arid regions. On the other hand, the analysis of spatial information affecting the presence of groundwater water needs powerful analytical methods which is able to interpret non-linear relationships, internal relationship and hidden information between affecting layers. In this regard, the current research is planned with the aim of comparative evaluation of the spatial changes of groundwater potential using Frequency Ratio (FR) and Weight of Evidence (WoE) models in the Esfarayen county, North Khorasan province. The factors of elevation above sea level, slope aspect, slope degree, land use, geology, topographic wetness index (TWI), plan curvature, profile curvature, distance from the river, drainage density, stream power index (SPI), distance from faults and fault density were prepared as effective layers in ArcGIS platform. Based on FR and WoE models, the relationship of spatial maps and groundwater potential was determined. Evaluation of prepared maps was done using relative operating curve (ROC). Based on the results, the Frequency Ratio model with Area Unver the Curve (AUC) value equal to 84.4 percent showed better performance than WoE model in the study area. Also, geological factors, land use, elevation above sea level and distance from the river were selected as the most important factors affecting the groundwater potential based on the results of the employed models. In addition, 12.5%, 28.45%, 34.75%, 19.3%, and 5% of the study area are located in the classes with very low, low, medium, high, and very high potential of groundwater based on the FR model, respectively. The results of the present research can be used in decision-making related to water resources management and locating areas prone to groundwater extraction or adopting optimal aquifer management strategies in Esfarayen county.