Karst aquifers represent, in large areas of the world, a strategic socio-economic and environmental resource providing drinkable water, a great biodiversity and amazing landscape. Ford and Williams (2007) estimated that about 25% of the world’s population depends on water coming from karst aquifers. For this reason, from the 70’s, several studies on European karst aquifers have been carried out, in order to create conceptual models and reproduce the hydrogeological behavior. The main issue of these models was the hydraulic role of the epikarst and the “duality” (Kiraly, 1994) of infiltration process (diffuse and rapid) , groundwater flow field (low velocity in the matrix and high in conduits) and discharge condition (low and constant and high and variable), which is a direct consequence of the heterogeneous structure of these aquifers. In Italy, only from the late 70’s hydrologists and hydrogeologists focused their studies on karst aquifers, for enhancing the protection, usability and exploitation of water resources. Different models were proposed for the Alpine (Civita et alii, 1991;​ Vigna, 2007) and the southern Apennines (Celico, 1978;​ Celico, 1983a;​ Celico, 1986;​ Celico et alii, 2001;​ Allocca et alii, 2007) karst aquifers.

The latter are currently important sources also for several bottling plants and feed different aqueduct systems. In order to avoid pollution and over-exploitation and for a fair management of the resource, the correct estimation at various space-time scales of groundwater recharge processes, corresponding to mean annual replenishment of aquifers by infiltration processes through the vadose zone, is a fundamental and challenging issue, also considering the climate variability affecting the area.

On a regional scale, the impact of long-term climate variability on groundwater recharge was analysed by using regional indexes of precipitations, temperature, effective precipitations and spring discharges. In order to calculate these climate indexes, times series of precipitations and air temperature were gathered, from 1921 to 2012, from 18 rain gauge stations and 9 thermometric stations, chosen in function of the continuity of functioning and spatial distribution over the territory. Moreover, for the same period, discharge data were collected from three karst springs. A strong correlation was found between the climate indexes and the pattern of the North Atlantic Oscillation Index, analysed using different statistical techniques (using smoothing numerical techniques, cross-correlation and Fourier analysis).

On a basin scale, an original Annual Groundwater Recharge Coefficient (AGRC) was created, by means of an integrated approach based on hydrogeological, hydrological, geomorphological, land use and soil cover analyses. The coefficient represent the ratio between net groundwater outflow and the precipitation minus actual evapotranspiration (P-ETR) for a karst aquifer. To estimate P-ETR mean value for each considered aquifer, was constructed a model for the whole southern Apennines, based on the relation with the altitude, the orographic barrier and the rain shadow effect. The AGRC values found for the sample aquifers ranges from 50% to 79% and are consistent with those estimated for karst aquifers in Europe from other authors. Moreover, considering the peculiar geomorphological feature of the karst aquifers, characterized by huge summit plateau and endorheic areas, namely by a total infiltration and no runoff, an additional coefficient was assessed in order to estimate the recharge for the slope areas only (AGRCS). Groundwater recharge in karst aquifers:​ estimation at different spatial and temporal scale and effect of climate variability

In addition, was studied the effect of the parameters affecting recharge on the estimated AGRC. A multiple linear regression between the AGRC, lithology and the summit plateau and endorheic areas was found. The latter is a valid tool for estimate recharge and, consequently, the mean annual runoff also for those aquifers characterized by a lack of spring discharge time series.

On a local scale, the groundwater recharge of a test perched karstic aquifer, belonging to the Mount Terminio hydrogeological structure was estimated. For such a purpose, an improvement of the Water Table Fluctuation (WTF) method, known as Episodic Master Recession (EMR) method (Nimmo et alii, 2015), was applied to estimate the Recharge to Precipitation Ratio (RPR) coefficient, which represents the amount of precipitation recharging groundwater. RPR values varies between 35% and 97%, with an annual annual RPR value (73%) resulted well matching with the AGRC estimated for that portion of Terminio karst aquifer (70%) and with the evapotranspiration value (30%) assessed through the use of Thornthwaite-Mather’s method. By a multiple regression analysis, RPR values are recognized as chiefly influenced, mainly, by the rainfall intensity and, secondly, by the antecedent soil water content.

The different methods used at different spatial and temporal scale represent an advancement of the understanding of the hydrological behavior of karst aquifers and in particular of the recharge process. For the first time, has been establish a link between a large-scale atmospheric cycle and the groundwater recharge of carbonate karst aquifers. Furthermore, has been carried out an original coefficient (AGRC) through a multidisciplinary approach that can substitute those heuristically calculated (Celico, 1986;​ Allocca et alii, 2007) in the water balance calculations. Moreover, the EMR method has been applied in a heterogeneous perched karstic aquifer of the southern Italy, whereas, until now, was used in few porous aquifers, with homogeneous hydraulic properties (Nimmo et alii, 2015).