Mapping the hazard of groundwater pollution by disjunctive kriging

The hazard of an event that is detrimental for a natural system can be combined with the vulnerability of that system to obtain the risk. In this paper an approach to assess the level of spatial hazard for groundwater quality based on Disjunctive Kriging (DK) is presented. The DK offers the possibility to evaluate, at a fixed time, the Conditional Probability (CP) for a target pollutant (i.e. a chemical parameter sampled in wells) to override a given threshold (i.e. concentration value) in any point of a study area. Varying t, different hazard maps can be plotted. A non parametrical statistics allows to evaluate the temporal trend of the CPs. The map of the trend index can be combined by the most recent map of the spatial hazard. The effective levels of the space-time hazard will result in a classification according to the rules coming from a tabular combination. As a case study, effective hazard levels maps of groundwater quality have been plotted for the aquifer of the plain of Modena, Northern Italy. Intensive agriculture and pig breeding take place in the study area along with industrial and civil settlements. The influence of the agriculture on groundwater results in high nitrate pollution that limit the use of such a resource for drinking water supply.