Extended Abstract
Introduction. In recent years, the population growth, the increase in irrigated land and economic development have caused the increase in the demand for groundwater resources all over the world. In arid and semi-arid regions where surface water does not have a significant amount due to low rainfall and high evapotranspiration, people lives mainly depend on groundwater. As a result, it is necessary to identify the groundwater potential areas and determine its recharge areas using accurate technologies. So, the aim of this research is to compare the method of multi- influencing factors with the fuzzy method for determining the potential of groundwater in a part of Kebar-Fordo watershed, Qom city, Iran.
Materials & Methods. For this purpose, a part of Kebar-Fordo watershed located in Qom province was selected. Six factors layer, viz. slope, annual rainfall, distance from river, geology, soil, and landuse were considered and classified based on groundwater potential susceptibility in different scales. Multi-influencing factor method can determine the groundwater potential of the region by assigning appropriate weight to different effective factors. In this approach, the layers were combined in Arc-GIS after determining the weight of the layers. In the fuzzy method, the layers of six factors were converted to fuzzy based on the linear function, and then the layers were incorporated using the gamma function. Finally, the statistics of observation points and accuracy index were used in order to evaluate the models,
Results & Discussion. The slope map represents that most part of the studied area (78.56%) has a "0-1" class while "1-3", "3-9" and "9-25" slope classes could be observed in 19.97, 1.29 and 0.18% of the total area, respectively. The soil texture has a significant effect on the infiltration and percolation of the surface water movement towards the groundwater. Therefore, in this research, the soil factor has been investigated as one of the input factors to the models. Soils with high permeability are more suitable for groundwater recharge and vice versa. The soil texture of the area consists of sandy loam, loam, sandy clay loam, and clay loam textures, which cover 3.73, 90.72, 0.23, and 5.32% of the total area, respectively, with a rank of four to one  for groundwater potential. In this study, geology map showed that Qft₂ formation has the largest area (88.98%) and Plc formation is in the second rank (4.9%). Qft₁, Qs.d and Mur units have an area of 2.22, 2.12 and 1.10% and the smallest area belongs to OMq formation (0.68%). Also, different types of land use in the study area were agriculture, garden, rangeland, bareland, and resendential area. The land use map showed that the largest area of this area was ariculture landuse (77.18%), while garden and rangeland covered 0.07 and 6.5% of the total area, respectively. Bareland and residential area comprise 2.94%, 13.31% of the total area, respectively. Among the different landuses, agriculture and residential area have the highest and lowest ranks in groundwater recharge. The rainfall map was categorized with four classes. The classes of 140-156, 156-168, 168-182, and 182-203 mm layers include 14.15, 48.92, 21.84 and 15.09% of the total area with the rank of one to four for groundwater recharge, respectively. The map of distance from the stream was divided into four categories: "0-659", "659-1480", "1480-2675" and "2675-4939" meters, which comprise 46.33%, 34.15%, 15.72% and 3.8% of the total area, respectively. In the method of multi influencing factor, the distance from the stream (8.33%) and the geological factor (25%) were the lowest and highest weights. In this regard, the factors of rainfall, slope, soil, landuse have 20.83%, 16.67%, 16.67% and 12.5% weights, respectively. Then, the groundwater potential map was prepared through overlaying in ArcGIS and the studied area was classified into suitable and unsuitable classes. The suitable class covers 75.15% of the studied area and the unsuitable class covers 24.85% of the total area. In the fuzzy method, the unsuitable class comprises 43.63% and suitable class covers 56.37% of the area. In order to evaluate the models, the statistics of the observation points were applied which the accuracy of the multi- influencing factor and fuzzy models was calculated as 71.42 and 78.57%, respectively.
Conclusion. Preparation of groundwater potential map is necessary to adopt management measures of rainfall storage and groundwater recharge in arid and semi-arid regions and it can be used for sustainable management of groundwater resources. The findings of this research revealed both model's accuracy in the studied area.