Stochastic modelling of solute mass discharge to identify potential source zones of groundwater diffuse pollution.

In heavily urbanised areas, groundwater diffuse pollution is recognised as one of the most insidious threats to groundwater quality. Diffuse pollution originates from multiple small sources releasing a low contaminant mass over a relatively large area; the lack of a defined plume in groundwater, the limited leaked mass, and the fact that leakage may have occurred in the past and be now ceased, make these sources difficult to locate and characterise. In addressing this environmental issue, an inverse approach based on the Null space Monte Carlo stochastic method has been applied in the framework of an innovative methodology with the aim to locate potential source areas distributed in a large (120 km2) urban area. To simplify the problem and better understand the limitations and effectiveness of the proposed methodology, the analysis has been performed using a groundwater model with fixed (i.e., determined by a previous calibration) hydraulic conductivity and flow boundary conditions. The only source of uncertainty considered in the study is the PCE mass discharge from all model cells of the topmost layer. After implementing and calibrating a deterministic solute transport model, multiple random realisations of mass discharge fields were generated, all of which are history-match constrained and hydrogeologically plausible. The obtained stochastic parameter sets were used to investigate the statistical distribution of the solute mass discharge and map the areas that are more likely to host unknown sources of PCE. Although the application of the NSMC stochastic method on the synthetic case study has provided promising results, it has also highlighted that multiple sources of uncertainty (e.g., continuity and duration of each source, attenuation processes) could adversely affect the reliability of the results in a real-world context, in which the effect of other uncertain parameters (hydraulic conductivity amongst all) would need to be considered in addition. This study offers new insights to the problem of aquifer diffuse pollution by providing key information on the potential source zones and on the areas that urgently need to be prioritised for further investigations.

Multi-aquifer susceptibility analyses for supporting groundwater management in urban areas.

In the densely urbanised Milan Metropolitan area (northern Italy), the long history of anthropogenic activities still exerts a significant pressure on groundwater resource. One of the most serious threats to the water quality of urban aquifers is attributed to diffuse contamination, which is caused by a series of unknown small sources (i.e., multiple point sources) distributed over large areas. In the study area and in many industrialised regions of the world, tetrachloroethylene [PCE], trichloroethylene [TCE] and hexavalent chromium [Cr(VI)] represent the common example of long-standing and persistent pollution in groundwater. In the Milan Metropolitan area, high levels of PCE + TCE and Cr(VI) were detected in the shallow aquifer as well as in the deep aquifer. To assess and map the shallow and deep aquifers susceptibility to PCE + TCE and Cr(VI) contamination at a regional scale, the Weights of Evidence modelling technique has been applied. This method has been used to objectively evaluate the spatial correlation between the high presence of these pollutants in each aquifer and hydrogeological and land use factors that can potentially influence the contamination. Moreover, the results allowed us to quantify on a large scale the effect that preferential flowpaths, due to both thickness variation in the aquitard and the areal density of multi aquifer wells, have in reducing the protection of the underlying deep aquifer. The end-products of the study constitute a key tool to be used by water-resource managers and decision-makers for the improvement of groundwater management and protection strategies.

PCE point source apportionment using a GIS-based statistical technique combined with stochastic modelling.

To meet the continuous growth of urbanised areas with the ever-increasing demand for safe water supplies, the implementation of new scientifically based methodologies can represent a key support for preventing groundwater quality deterioration. In this study, a new combined approach based on the application of the Weights of Evidence and the Null-Space Monte Carlo particle back-tracking methods was set up to assess tetrachloroethylene (PCE) contamination due to Point Sources in the densely urbanised north-eastern sector of the Milano FUA (Functional Urban Area). This combined approach offers the advantage of further enhancing the power of each individual technique by integrating both the advective transport mechanism, neglected by the Weights of Evidence, and the influence of specific factors, such as the land use variation, not considered by the Null-Space Monte Carlo particle tracking. To accurately test and explore the performance of this new approach, the analysis was carried out based on the simulation of synthetic PCE plumes using a groundwater numerical model already implemented in a previous study. The Weights of Evidence method revealed that the areas characterised by a groundwater depth lower than 17 m, a groundwater velocity higher than 2.6 × 10-6 m/s, a recharge higher than 0.26 m/y and a significant variation of the industrial activities extent are the most susceptible to groundwater pollution. The Null-Space Monte Carlo particle back-tracking has proved to be effective in delineating the potential source zones and contaminant travel path. The proposed approach can offer additional insights for the protection of groundwater resource. The end-product provides crucial information on the zones that require to be prioritised for investigations and can be easily understood by non-expert decision-makers constituting an advanced tool for enhancing groundwater protection strategies.