Multivariate data analysis and numerical modeling: a combined approach to assess the Groundwater contamination in Milan Functional Urban Area

The Italian law D.Lgs.152/2006 adopted the EU Water Framework Directive principles and entrusted the Regions with the task to identify the areas affected by groundwater diffuse pollution and to assess their contamination level. The proposed methodology (the Whole work has been supported by Regione Lombardia and ARPA Lombardia within the PLUMES Project) combines the numerical model approach, which defines the areas interested by a plume transit, with the statistical approach, in order to identify both point-source contamination (i.e. hot spots) and multiple point contamination that determine a status of diffuse contamination. The methodology was applied to the Functional Urban Area of Milan, where, at least since 40 years, chlorinated aliphatic hydrocarbons (TCE, PCE, TCM) have been the main groundwater contaminants in the unconfined and confined Aquifers and they have been constantly monitored by public authorities. The study was divided into 2 different steps. In the former, the Cluster Analysis (CA, Vega et al., 1998; Otto, 1998) allowed to identify, for each contaminant, the hot spots that have been used as sources in the numerical transport model (MT3DMS, Zheng & Wang, 1999 and MODFLOW-2000, Harbaugh et al., 2000). The contaminant transport model, calibrated on the concentration values of 2014, gave important information about the influence of the plumes on the groundwater chemical status observed in the monitoring wells network. In the latter step, the monitoring wells located inside the plume areas were removed from the ARPA Lombardia contaminant concentration value dataset with the aim to keep just the concentrations representing the multiple-point diffuse contamination component. The new dataset was then used for the geostatistical analysis (Inverse Distance Weighted) in order to map the diffuse contamination, not directly influenced by hot spots. Lastly, the work defined reference values of diffuse contamination levels identified in the maps. To sum up, the results of the work provide a better understanding of the relationship between hotspot sources, multiple sources and diffuse pollution in the FU