Some limitations in using 222Rn to assess river–groundwater interactions: the case of Castel di Sangro alluvial plain (central Italy)

222Rn was used to assess river-groundwater interactions within Castel di Sangro alluvial aquifer (Italy). The effectiveness of results obtained through this indicator was verified by also analyzing δ18O, major ions and temperature in both surface and groundwater, and carrying out piezometric head monitoring and discharge measurements. Hydrogeological investigations suggested that the river infiltrates into the aquifer in the south-eastern aquifer portion, while groundwater discharges into the river in the north-eastern portion. The latter phenomenon is supported by 222Rn data. Nevertheless, flow-through conditions cause the modelled discharge along this river reach, estimated by 222Rn data in a degassing-corrected two-component mixing model, to be greater than the measured discharge. Concerning river infiltration into the aquifer, δ18O, major ions and temperature data show that the river contribution is negligible in terms of aquifer recharge. Thus, the observed increase in 222Rn concentration in that portion of the aquifer is due to the enrichment process caused by infiltration of rainwater (222Rn free) which flows from the local divide area. Hence, in the study site, the use of only 222Rn to predict river-groundwater interactions causes some estimation inaccuracies and it must be coupled with other hydrochemical and hydrogeological parameters to gain a thorough understanding of such interactions