Potential aquifer characterization through the integration of multivariate geophysical datasets

The work developed in this thesis has the aim to provide a methodology for geophysical characterization of an aquifer. Electrical resistivity tomography and seismic P-waves refraction/reflection were applied to explore the subsurface. Thanks to their sensitivity to factors as mineralogy, porosity, texture and pore content, electrical resistivity and acoustic waves can depict subsurface changes within an aquifer. The coupled use of both geophysical parameters allow to obtain complementary informations about the subsurface, reducing the uncertainty with respect to the univariate case. Two different approaches were proposed to characterize the shallow part and the deep of the subsurface. Two post-processing integration methods were proposed, where the shallower is based on electrical resistivity and P-waves seismic velocity from tomographic techniques and the deeper on seismic attributes datasets from seismic reflection. The results are a shallow unique model characterizing the subsurface with geophysical facies corresponding to different lithological conditions and a high-resolution large-scale image depicting main structural and stratigraphic changes for the aquifer characterization. The two approaches are based on clustering techniques: k-means and Silhouette were applied to find out the optimal number of geophysical facies from tomographic dataset and Self-organizing maps were used to highlight pattern in the seismic attributes dataset. Integration methods were applied in two different geological and hydrogeological environments: an aquifer in intra-Apennine basin (Laterina, Italy) and a complex hydrothermal system into the Solfatara volcano (Naples, Italy). In the case of Laterina, electrical resistivity tomography and seismic reflection/refraction surveys were carried out to characterize shallow unconfined aquifers in alluvial deposits and a deeper confined aquifer in rock. The tomographic integration revealed six geophysical facies consistent with stratigraphic log. Moreover allowed to image the 2D disposition of main aquifers and aquitards. The integration of seismic attributes provided a large-scale image of the highlighting shallow alluvial recent and Villafranchian fluvial-lacustrine deposits above the Monte Cervarola Unit (sandstone/marls). The integration processes were applied at the case of Solfatara to highlight the shallower fluids paths, and the main structural paths of uprising for fluids within the caldera. The tomographic images provided three different geophysical facies, one consistent with poorly consolidated sediments, and two high velocity low and high resistivity, respectively associable with liquid dominated and gas dominated media. The integrated model obtained by SOM provided a image of the caldera with sub vertical discontinuities interpreted as fracture or ring fault, main paths for fluids circulation. Moreover, several sub-horizontal bright spot were depicted within the caldera, consistent with vapor-saturated media.