Recovery from groundwater extraction in a small catchment area with crystalline bedrock and thin soil cover in Sweden

An experiment has been in progress since 1997 in a small catchment area (28,000 m2) with crystalline bedrock and thin soil cover to study the conceivable impact on groundwater conditions of tunneling and the use of groundwater. The impact on hydrology and hydrochemistry from intensive extraction of groundwater at a depth of 50 m in the bedrock has been studied at Lake Gårdsjön in Sweden. The catchment area was first monitored under pristine conditions, followed by four and a half years of extraction and then a recovery phase. The geological conditions result in a low buffer capacity and high sensitivity to acidification. During the period of extraction, the surface runoff decreased by approximately 50% compared to a nearby reference area. The groundwater extraction caused increased fluctuation in groundwater levels in a wetland, which in turn caused oxidation of reduced sulfur to sulfate. The sulfate concentrations increased almost 100-fold in some instances, causing a lowering of the pH by one unit in shallow groundwater. Since extraction of the groundwater was discontinued, the pH has gradually risen and the sulfate concentrations have decreased. However, the concentration of sulfate in groundwater in the wetland has remained stable at approximately double the pre-experiment levels. Magnesium concentrations were lower after the experiment, caused by exhaustion of the magnesium pool in the wetland through acidification. The extraction of water from the bedrock shortened the retention times and increased the recharge of groundwater in the bedrock. After extraction was terminated, the groundwater levels in the boreholes recovered within a month to levels similar to those before extraction. The hydrochemistry of the bedrock groundwater, which was strongly affected by the hydrochemistry of shallow groundwater during the experiment, has also gradually begun to regain its pre-extraction signature. However, the surface runoff has remained low during the first 2 years of recovery, at about 60% of the volume compared to the unaffected catchment area. This could be explained by delayed recovery in resaturation of the shallow rock that was unsaturated during the experiment.