Using Different Thresholds in Assessing Groundwater Vulnerability through Statistical Methods

Statistical techniques are widely used in environmental studies to evaluate natural hazards and support decisions about environmental planning and management. In groundwater vulnerability, the determination of a threshold concentration to discriminate between impacted or non impacted wells represents a key issue in the application of these techniques. In this study, the effects on groundwater vulnerability assessment due to the use of different threshold values for modeling have been evaluated using the Weights of Evidence statistical method for a porous shallow unconfined aquifer in Northern Italy in a 2000 km2 area. A guided procedure has been proposed and applied to select three vulnerability maps, one for each different threshold values of nitrate concentration in groundwater used to subdivide the monitoring wells. Some important aspects have been considered in the presented approach: a) dealing with problems of conditional independence; b) statistical significance and relative importance of classes of the explanatory variables c) reclassification of groundwater vulnerability maps; d) evaluation of the reliability of the reclassified maps. The procedure allowed to effectively compare results obtained for each threshold, highlighting advantages and drawbacks in their use, and to test the general robustness of the method for groundwater vulnerability assessment.