Climate change and future groundwater level projections for the Salento Aquifer (Italy)

Groundwater is a vital source of fresh water in many areas of the  Mediterranean Basin and it will become increasingly important due to climate  change and the resulting intensification in water scarcity. The Salento peninsula  (Puglia region) in southern Italy is no exception, where groundwater is the main  source mostly for drinking and irrigation purposes. In fact, the karst nature of the  Salento territory allows the presence of groundwater resources rather than  surface water resources. The recharge conditions of this aquifer are very complex  since its geological and hydrogeological structures are complex. In addition to  this, the scarcity of data prevents to understand all the phenomena as a whole  and to setup integrated surface/groundwater models that could be useful for  investigating the impacts of climate change on the Salento groundwater  resources. However, for the Salento area, a direct correlation between  hydrological variables (precipitation and temperature) and groundwater levels  (GWLs) has been shown. In particular, despite the complexity of the aquifer, it is  possible to define potential linear relationships between observed meteorological indices, such as the Standardized Precipitation Index (SPI) and the Standardized  Precipitation Evapotranspiration Index (SPEI), and GWLs collected in monitoring  wells. In most cases, there are high correlation coefficients between SPI/SPEI and  GWLs for long accumulation periods of the climate variables, in the order of 18  months. This, for the aquifer studied, reflects the attenuated response of  groundwater to precipitation/precipitation-evapotranspiration signals. In this  work, we use future projections of SPIs and SPEIs for the Salento area to estimate  the possible impacts of climate change on the groundwater resource. In  particular, for this study, we computed SPI and SPEI projections, until the end of  this century, from future precipitation and temperature data of an ensemble of  Regional Climate Models (RCMs) of the EURO-CORDEX initiative, under two climate scenarios, namely the RCP4.5 and the RCP8.5. Prior to their use, the raw  RCM data were downscaled/bias corrected to properly represent the observed  climate in the study area. Under the hypothesis that the observed linear  relationships between meteorological indices and GWLs remain valid over time,  we obtained future projections of GWLs from the projections of SPIs/SPEIs. The  analysis of the future GWLs has been conducted at short-, medium- and long term. The use of an ensemble of climate models allows in evaluating the  uncertainty of the results. As long as a high correlation between meteorological  indices and GWLs exists, the approach used in this work can be easily applicable  to other study areas.