Testing the usefulness of 222 Rn to complement conventional hydrochemical data to trace groundwater provenance in complex multi-layered aquifers. Application to the Úbeda aquifer system (Jaén, SE Spain).

The Úbeda aquifer system is a multi-layered aquifer intensively exploited for irrigation. It covers 1100km2 and consists of piled up sedimentary aquifer and aquitard layers from Triassic sandstones and clays at the bottom, to Jurassic carbonates (main exploited layer) in the middle, and Miocene sandstones and marls at the top. Flow network modification by intense exploitation and the existence of deep faults favour vertical mixing of waters from different layers and with distinct chemical composition. This induces quality loss and fosters risk of quantity restrictions. To support future groundwater abstraction management, a hydrogeochemical (major and some minor solutes) and isotopic (222 Rn) study was performed to identify the chemical signatures of the different layers and their mixing proportions in mixed samples. The study of 134 groundwater samples allowed a preliminary identification of hydrochemical signatures and mixtures, but the existence of reducing conditions in the most exploited sector prevents the utility of sulphate as a tracer of Triassic groundwater in the Jurassic boreholes. The potential of222 Rn to establish isotopic signatures and to trace groundwater provenance in mixtures was tested.222 Rn was measured in 48 samples from springs and boreholes in most aquifer layers. At first, clear correlations were observed between222 Rn, Cl and SO4 in groundwater. Afterwards, very good correlations were observed between222 Rn and the chemical facies of the different layers established with End Member Mixing Analysis (EMMA). Using222 Rn as part of the signatures, EMMA helped to identify end-member samples, and to quantify the mixing proportions of water from the Triassic and the Deep Miocene layers in groundwater pumped by deep agricultural wells screened in the Jurassic. The incorporation of222 Rn to the study also allowed identifying the impact of irrigation returns through the association of moderate NO3 , Cl, and Br contents with very low222 Rn activities.