Estimation of stream flow depletion and uncertainty from discharge measurements in a small alluvial stream

Hydrographs were recorded at three discharge stations at a small alluvial stream during the summers of 1997 and 1998. A method of analysis was set up so that the temporal variations in discharge resulting from natural hydrological processes can be distinguished from the influence from ground water being periodically abstracted approximately 60 m from the stream. Thus, evapotranspiration from the riparian zone resulted in diurnal variations in streamflow with maximum amplitude of 3–5 l/s, whereas heavy rainfall resulted in intense short-term surface/subsurface flow from the riparian zone. The magnitude of peakflow was of the order of 2–3 times baseflow and such events typically disturbed the hydrograph for one to three days. Baseflow increased from about 35 l/s to about 70 l/s over the studied reach, but when ground water was abstracted at rates of about 15 l/s this resulted in a reduction in discharge. Within 4–8 days the reduction stabilized at about 4 l/s and at 5–7 l/s, respectively, at discharge stations 140 and 350 m downstream of the well site. Predictions made by the analytical depletion model by [Ground Water, 37 (1999) 98] using the values of transmissivity and conductance of the streambed estimated by [Ground Water, 40 (2002) 437] by drawdown analysis compare reasonably well to the observed reduction of streamflow.