Changes in groundwater reserves and radiocarbon and chloride content due to a wet period intercalated in an arid climate sequence in a large unconfined aquifer

Abstract The concentration of atmospheric tracers in groundwater samples collected from springs and deep wells is, in most cases, the result of a mixture of waters with a wide range of residence times in the ground. Such is the case of an unconfined aquifer recharged over all its surface area. Concentrations greatly differ from the homogeneous residence time case. Data interpretation relies on knowledge of the groundwater flow pattern. To study relatively large systems, the conservative ion chloride and the decaying radiocarbon ( 14 C) are considered. Radiocarbon ( 14 C) activity in groundwater, after correction to discount the non-biogenic contribution, is often taken as an indication of water age, while chloride can be used to quantify recharge. In both cases, the observed tracer content in groundwater is an average value over a wide range which is related to water renewal time in the ground. This is shown considering an unconfined aquifer recharged all over its area under arid conditions, in which a period of greater recharge happened some millennia ago. The mathematical solution is given. As the solution cannot be made general, to show and discuss the changes in water reserve and in chloride and radiocarbon concentration (apparent ages), two scenarios are worked out, which are loosely related to current conditions in Northern Chile. It is shown that tracer concentration and the estimated water age are not directly related to the time since recharge took place. The existence of a previous wetter-than-present period has an important and lasting effect on current aquifer water reserves and chloride concentration, although the effect on radiocarbon activity is less pronounced. Chloride concentrations are smaller than in current recharge and apparent 14 C ages do not coincide with the timing, duration and characteristics of the wet period, except in the case in which recharge before and after the wet period is negligible and dead aquifer reserves are non-significant. The use of chloride concentration in springs as a proxy of chloride concentration in recharge to estimate recharge from atmospheric deposition leads to recharge value larger than the real one and it approaches the wet period recharge. Drawing inferences about radiocarbon data and recharge by the chloride balance method has rarely been taken into account before. It is important to consider the variable aquifer groundwater reserve. Current recharge estimation can be improved by careful selection of groundwater samples, supported by tritium and radiocarbon measurements.