Applying Spatial Statistics to Model the Patterns Governing Distribution of Wells (Case Study: Mahidasht Watershed, Kermanshah)

Despite the many capabilities of spatial statistics in modeling and explaining various environmental features and identifying the patterns governing their distribution, the previous studies have considerably neglected these capabilities, especially in the field of water studies. The present study aims to model the patterns governing the distribution of water wells in the Mahidasht watershed in Kermanshah province using spatial statistics. The bases used to identify these patterns are the "Locations" and "Attributes" of the wells, which are performed using five common techniques in spatial statistics, including: Average Nearest neighbor, Getis-Ord General G, Spatial Autocorrelation (Moran I), Cluster and outlier analysis, and Hot spot analysis. Required data were obtained from the Kermanshah Regional Water Company. The results of the present study showed that the pattern governing the dispersion of agricultural wells and the pattern governing the dispersion of high values (high depth of wells) in the Mahidasht watershed do not confirm the null hypothesis and follow the cluster pattern. The general pattern of dispersion of wells in the study area since 1950s to 2000s has changed from a random pattern to the cluster one. Cluster and outlier analyses as well as hotspot analysis indicated that the northern and northeastern parts of the watershed are critical in terms of uncontrolled abstraction of groundwater resources.