Recharge estimation by stage-discharge interpolation of springflows from cross-correlated well measurements

For shallow phreatic aquifers, from which springflow is the predominant source of discharge, recharge volumes may be approximated as the time integral of springflow time series. However, field measurement of flows from numerous springs,,vith sufficient frequency to accurately estimate this time integral, is frequently impractical. A technique for interpolation of continuous springflow time series is presented, using cross-correlation between flows and continuous head measurements in associated wells, Suitable wells must lie within the recharge catchment of the spring in question and display negligible phase lag between well and spring response to recharge. Brief (<1 year) periods of calibration may be sufficient to establish calibration upon which discharge interpolation from well data may be based. A 7 m thick highly heterogeneous mine-waste aquifer, with springflows highly sensitive to recharge events, is examined as a test case. Over a 40 day recharge period, three well/spring pairs yielded acceptable cross-correlation across four recharge events, Head from one of these wells was compared to a second nearby spring and found to show poor correlation, attributed to its location outside of that spring's catchment. Calibration statistics were found to be robust when the calibration dataset for one pair was reduced in size. Measured flows from one of the correlated springs also displayed correlation with summed discharge from the aquifer at all measured springs, based on concurrent flow measurements over time. Results suggest that continuous measurements of head from a small number of critically selected wells may allow accurate estimates of transient spring discharge rates following a relatively brief period of flow-stage calibration. The success of the method may depend on specific characteristics of aquifers to which it is applied, such as size of spring catchments, aquifer conductance, aquifer heterogeneity, and spatial distribution of recharge.