Hydrogeological and hydrogeochemical monitoring to identify “hydrosensitive sites” in central-southern italy

In the last decades, groundwater and gas responses to seismicity have been widely documented in tectonically active regions. Many studies have highlighted the sensitivity of fluid behaviour related to the seismic activity both in terms of hydrogeological and hydrogeochemical anomalies. For instance, variations, including changes in spring discharge, groundwater level, geochemical content, isotope composition, dissolved and free gases have been observed in sensitive monitoring sites at different distances from epicentres. Through the years, various mechanisms have been proposed in the literature to explain groundwater level and discharge changes such as pore-pressure response to crustal elastic strain, permeability changes caused by seismic waves, and fluid migration along dilatant cracks or deep crustal fractures. Besides, variations in geochemistry and isotopic signature of groundwater have been also defined as the results of the following processes: rock weathering enhancement in new rupturing, mixing of waters from different aquifers, deep and hydrothermal fluid upwelling. Therefore, monitoring deep fluids contribution to shallow aquifer systems has been abundantly used for studies devoted to the understanding of the seismicity-groundwater relationship. Besides, defining fluids origin and possible water-gas-rock interaction processes during their upward migration can provide new constraints aimed at the comprehension of the seismogenic processes. In many countries all over the world, the set-up of hydrogeochemical networks has been already established or is currently in progress in order to identify responses induced by earthquakes and ascending deep fluids. Scientific efforts are now moving in this direction to obtain more observations and to build up reliable hydrogeochemical models associated with different geo-tectonic contexts. The aim of this PhD research activity is to take a step forward for the improvement of the comprehension of the aspect named Hydrosensitivity which is the “ability of a hydrogeological system to respond to external perturbations (e.g., earthquakes, groundwater mixing, interaction with deep fluids, etc.) to be further correlated with crustal deformation processes and consequently with micro- and macro-seismicity”. In detail, this study has the dual purpose of (i) identifying potential earthquake-induced groundwater and gas changes (i.e., direct groundwater-seismicity relationship), and (ii) analysing in-depth mixing of deep fluids in shallow regional aquifer systems in active seismic areas. To achieve these objectives, by considering hydrogeological and seismological features, monitoring sites in the central-southern Apennines were selected (i.e., Matese, Contursi, Sulmona areas, and San Vittorino plain) for monitoring geofluids as markers of active seismogenic processes. Two different approaches were adopted (i) long-term hydrogeological and hydrogeochemical monitoring was carried out with the seismic characterization of the study area for the detection of groundwater and gas variations related to the seismicity, (ii) complete hydrogeochemical characterization of groundwater to define shallow and deep fluids flowpaths and their consequent mixing was realized through several elaborations. Overall, the results of this study confirm the great potential of having continuous hydrogeological and hydrogeochemical monitoring to identify variations induced by seismicity also related to the occurrence of small-intermediate earthquake magnitude, at least in the areas where deep fluids contribution to groundwater is evident. In fact, despite the absence of strong seismic events during the monitoring period, significant presismic hydrogeochemical signals, possibly linked to seismic activity, were identified at Grassano spring (Matese area). In particular, observed changes were referred to the Benevento seismic swarm, whose strongest event was recorded in San Leucio del Sannio on December 16th, 2019 with a magnitude of 3.9. In detail, starting from three months before the mainshock, an increase in dissolved CO2 and a lowering of pH were recorded. Furthermore, in conjunction with the seismic sequence, anomalous values of major ions (Ca2+, Na+, HCO3-) were detected, with a slight increase of the electrical conductivity. These hydrogeochemical variations were attributed to presismic crustal dilation processes that could allow the upwelling of deep CO2 with the Grassano groundwater, enhancing the ion solubility. The results highlighted the possibility of having potential presismic signals similar to those proposed in literature for the stronger seismic sequence of 2016-2017 Amatrice-Norcia, even for small-intermediate magnitude earthquakes. However, the groundwater-seismicity relationship was deepened not only by processing of data actively collected during the three years of research but also by analyzing previous data from the Matese and Sulmona area. Gas-geochemical monitoring of radon gas showed significant increases in dissolved gas activity in two springs before three seismic events (Balsorano earthquake Mw 4.4, 2019; L’Aquila earthquake Mw 3.8, 2018; San Leucio del Sannio earthquake Mw 3.9, 2019). The sensitive behaviour of Rn in relation to crustal deformation processes can be explained as the result of the dilation of the rocks-bearing aquifer which is expected to drive some changes in the chemical content of groundwater. Furthermore, the analysis of the piezometric data measured at the PF60.3 monitoring well in the Sulmona area allowed to distinguish impulsive post-seismic fluctuations in the piezometric level due to the passage of Rayleigh seismic waves deriving from distant earthquakes (up to 18.000 km). The hydrogeological responses highlighted the potential sensitivity of the fractured aquifers to the stress variation. Additionally, this study highlights once again that the acquisition of long-term hydrogeochemical time series (both discrete and continuous) is a powerful tool for analysing mixing and related processes in active seismic areas, where deep-seated normal faults could represent preferential pathways for the upward migration of fluids due to their high permeability. In detail, in the Contursi area, longer and more complex flowpaths were defined for thermal groundwater mixed with highly mineralized rising deep fluids, while local groundwater flowpath were supposed for cold groundwater fed by regional carbonate aquifers. Furthermore, the hydrogeochemical results are in line with recent geophysical studies, conducted in this sector of the Apennines, which reported the presence of a large volume of deep CO2-rich fluids below the study area. Since previous studies recognized the presence of deep overpressure fluids as a mechanism for triggering earthquakes, a long-term, high-frequency hydrogeochemical multiparametric monitoring could be a key to recognize significant variations in the mixing ratio between the shallow end-member and the deep one when earthquakes or energetic seismic sequences take place. Finally, in the San Vittorino plain, the investigation of groundwater trace elements occurrence clarified the water-rock interaction processes, supporting the conceptual model based on the localized uprising of deep fluids which matched the distribution of high-angle faults. The analysis of gas isotopes confirmed the presence of a deep contribution with possible traces of mantle-derived helium. Through the geothermal calculations, fluids depths were estimated where Apennine earthquakes usually occur. Also in this case, knowing the degree of mixing of the different sources, modulated by the fault activity, is essential in order to make considerations and processing relating to crustal deformations and seismicity.