Groundwater Recharge in Crystalline Bedrock : Processes, Estimation, and Modelling

The groundwater level in a fractured rock aquifer in Sweden was found to respond quickly to rainfall, although the bedrock was covered by 10-m-thick tilt soil. A considerable portion of the response was caused by surface loading, i.e., by the weight increase of the soil due to the addition of water from precipitation, whereas the rest reflected recharge. The hypothesis that the bedrock aquifer was recharged by vertical flow from groundwater in the overlying soil was tested with a simple recharge model, in which the bedrock-groundwater levels were simulated from the soil-groundwater and estimated surface-loading variation. The model had three parameters: the ratio between the equivalent vertical hydraulic conductivity governing the recharge and the storage coefficient of the bedrock reservoir, the recession coefficient for the bedrock-groundwater level, and the bedrock-groundwater level at which the outflow ceases. The model could be reasonably well calibrated and validated to head observations in one of two boreholes. The fit to the seasonal variation was similar when calibrating the model with or without surface loading, but surface loading had to be included to property simulate individual recharge events. The relative temporal variation in the fluxes could be determined by the calibration. The variation in the recharge was from -10% to +25% in relation to the mean flux. The variation in the discharge was only 1%. By applying a storage coefficient of the reservoir of 5 x 10(-4), the simulated mean recharge was about 20 mm yr(-1). The results support the hypothesis that the bedrock-groundwater at the site is fed by local recharge from the overlying soil aquifer.