Interazione tra acquifero superficiale e profondo nella Piana di Tivoli (Roma): Approccio multi-isotopico e modello numerico geochimico

Interaction between shallow and deep groundwater flow systems has been investigated in the Tivoli Plain aquifer system (Rome, Central Italy). During the last decade an intense activity in the travertine quarries in the Acque Albule Basin has caused a significant drop in the water table of the shallow travertine aquifer. As a consequence, subsidence and high instability risk are affecting buildings in this area constructed on top of Holocene sediments composed of mainly silty clay with high level of organic content, which are underlying by travertine deposits. A multi-isotope approach was used to have a better understanding of interactions between shallow and deep aquifers and to improve the knowledge of the hydrogeological conceptual model, which has implication for groundwater management in the Tivoli plain. Environmental isotopes are largely used for investigation of water origin, residence time and flowpaths (Kendall et al., 1998; Coplen et al., 1999). They are also useful for a better understanding of chemical reactions during water-rock interaction. A combined hydrogeologic and isotopic investigation using chemical and isotopic tracers such as SO4/Cl, δ18O, δ 2H, 87Sr/86Sr, δ34S, and δ13C was carried out in order to determine the sources of water recharge to the aquifer, the origin of solutes, and the mixing processes, in the Tivoli Plain, a Quaternary basin filled by travertines (Faccenna et al., 2008). The study area is located 30 km East of Rome. The recharge areas for the shallow groundwater in the travertine aquifer are supposed to be the carbonate ridges of Lucretili and Tiburtini mountains (Capelli et al., 2005: Petitta et al., 2010). The travertine aquifer also receives a contribution of mineralized fluids from a deep aquifer contained in the buried meso-cenozoic carbonates, which are separated from the shallow aquifer by low-permeability volcanic and clayed deposits. Representative samples of the water cycle in Tivoli Plain, which included springs, lakes, deep groundwater and water from the quarries were sampled for chemical and isotope analysis. Base-flow springs are generally saturated or oversaturated with respect to calcite, which explains the travertine formation (Minissale et al., 2002). A large number of samples including groundwater and surface water were collected in the Acque Albule Basin, while other samples come from the recharge area (S4, S5, S6, P6) (Fig. 3.2). S2 is a spring located out of the Basin, which is directly fed by a deep contribution from buried carbonate bedrock. Major ion chemistry data showed a groundwater stratification in the travertine aquifer, associated with mixing of the shallow groundwater with discharge mineralized fluids from the deep aquifer, partially enhanced by increasing pumping in the quarries. Results indicate that the hydrochemistry of groundwater in Tivoli Plain and adjacent recharge areas is characterized by a mixing among three end-members: A. groundwater of recharge area, B. groundwater of the shallow travertine aquifer (Acque Albule Basin), C. groundwater of deep carbonate aquifer. The end-members are represented by three different geochemical facies: Facies A: Ca – HCO3 type groundwater: TDS (0-0,8 g L-1); SO4 (0-250 mg L-1); DIC (0-7 mmol kg-1); EC (0-2 mS cm-1). Facies B: Ca – HCO3–SO4 type groundwater: TDS (0,8-2,4 g L-1); SO4 (250-800 mg L-1); DIC (7-16 mmol kg-1); EC (2-3,5 mS cm-1). Facies C: Ca-Mg – HCO3-SO4 type groundwater: TDS (2,4-3,6 g L-1); SO4 (800-1200 mg L-1); DIC (16-18 mmol kg-1); EC (3,5-4,5 mS cm-1). A multi-isotope approach (18O, 2H in water, 34S and 18O in sulphate, 13C in DIC and 87Sr/86Sr ratios) has been adopted in the study to obtain a better understanding of interactions between shallow and deep groundwater. The stable isotope data, collected in rain stations at different altitude and in groundwater, suggest the existence of different flowpaths and mixing of shallow groundwater associated with recharge in the Tivoli Plain. Based on seasonal changes in 18O and 2H, the recharge contribution coming from the carbonate ridges to deep groundwater has also been documented. The 13C data in DIC show a wide range in 13C values that varies between -12.3‰ and +8.6‰. The more depleted 13C values are considered representatives of the recharge area, where a input of soil CO2 occurs during rainfall infiltration mixing with DIC from dissolution of carbonates. Samples from Acque Albule Basin show values between +0.4‰ and +8.6‰, where an input of 13C enriched CO2 is associated with a deep contribution of hydrothermal fluids from the buried carbonate aquifer. The correlations in chapter 4 show two separated sources for DIC in the water samples, with some samples (P5, S6, C4) placed in intermediate position, justified by the influence by mixing processes. The 34S and 18O data in sulphate also highlight the existence of two different sources for dissolved sulphates: the groundwater collected in Acque Albule Basin have sulphates which can be associated to the Triassic evaporites of the deep aquifer; otherwise, sulphates of secondary origin from the shallow aquifer characterize samples collected in the recharge area. The positive values of 34S (> 10‰) may exclude sulphate reduction as main process in sulphate contribution, especially because it could not explain the high sulphate concentration of the B-C facies. A possible relationship between dissolved sulphates and the occurrence of H2S uprising fluids in the shallow aquifer can be discarded. Finally, the 87Sr/86Sr data with values ranging between 0,7076 and 0,7082 confirm that the contribution of dissolved solutes is associated with two sources: marine carbonates from the deep aquifer (groundwater influenced by deep flowpaths); continental and volcanic deposits in case of the shallow aquifer (groundwater having not interaction with deep flowpaths). An Inverse Model carried out with Phreeqc 2.16 by Parkhurst & Appelo (1999) has developed a theoretical geochemical evolution with water-rock interaction processes. According to an inverse mixing model, it is possible to conclude that both dissolution/precipitation and ion exchange processes are the key of geochemical evolution along groundwater flowpath, confirmed also by the calculated mixing between deep and shallow aquifers. The chemical and isotope tracers provided information for distinguishing different sources of dissolved salts and different groundwater circulation in the Tivoli Plain. The results of this study have improved the hydrogeological conceptual model, which can be summarized as follows: • the Ca-HCO3 groundwater type represents a flow system fed directly by meteoric water in carbonate ridges of Lucretili and Tiburtini mountains, surrounding the Tivoli Plain. The flow system is subdivided in a shallower one, that fills directly the travertine aquifer of the Acque Albule Basin, and in a deeper one, circulating in the buried carbonate bedrock; • the Ca-Mg–HCO3-SO4 groundwater type represents a deeper circulation having a contribution of high salinity fluids, uprising from the deep carbonate aquifer, probably related to the Colli Albani volcanic district. Mixing processes, which characterize the travertine shallow aquifer have been recognized in several water samples, especially inside the quarries area. In this area the mixing between the two components is widely enhanced by the recent occurrence of intense pumping activity (Prestininzi, 2008). The chemistry of samples in this area corresponds to a Ca–HCO3–SO4 groundwater type. Deep saline fluids rise and mix with recharge water in the shallow aquifer, evolving across dissolution/precipitation and ion exchange processes. The 18O, 2H and 87Sr/86Sr isotope values confirmed the meteoric origin of the groundwater and the different flowpaths influencing the hydrochemistry composition of groundwater in Tivoli Plain. The existence of two different sources of groundwater is supported by the results of 34S data in sulphates and 13C data in DIC. Both these two tracers support the existence of mixing in the shallow aquifer, showing intermediate values in the samples which are characterized by relative lower salinity.