Environmental indexes and aquifer heterogeneity

The development of reliable methodologies to characterize aquifer heterogeneities is a key issue for the management and safety of groundwater resources. The effective reconstruction of the heterogeneity pattern is important to predict solute characteristics as well as spatial spreading, mixing and residence time of pollutants. So hydrogeological models often involve a geostatistical treatment of many data to address aquifer uncertainties. However, there may be wide areas where data are unavailable or insufficient for such approaches. This study shows that in this case the variability of hydrodynamic dispersion parameters can be related to a set of groundwater compositional indicators. At the same time, this analysis provides the identification of homogeneous sectors and suggests the more effective working scale, especially when flow systems extend across different aquifer types and are characterized by high anthropic impact. With regard to complex alluvial aquifer systems of Southern Italy, cluster analysis can support the definition of homogeneous groundwater groups taking into account the influence of natural mixing phenomena and anthropogenic pressures. The present paper deals with an explanatory case study. A set of 77 parameters was processed, including the most common heavy metals. It was obtained from groundwater samples of a pyroclastic alluvial aquifer, about 53 km2 in area. This chemical-physical characterization, resulting from samples collected on July 2010 (29 wells and 1 spring) and on December 2014 ((23 wells and 2 springs), was part of a more complete hydrogeological monitoring activity including monthly measurements of groundwater level, electrical conductivity, pH and temperature in 76 wells and 4 springs for about two years. In order to identify homogeneous groundwater groups a K-means clustering technique is applied to the most significant water composition indexes. In this regard an important pre-processing step to avoid suboptimal solutions is represented by the identification of the end member for the bedrock as well as alluvial plain aquifer. Identification of the groundwater end-members and the results of cluster analysis will be discussed in detail, pointing out the potentialities of the proposed approach in the assessment of groundwater mixing and flow