Discharge of groundwater into lakes (lacustrine groundwater discharge, LGD) can play a major role in water balances of lakes. Unfortunately, studies often neglect this input path because of methodological difficulties in its determination. Direct measurements of LGD are labor-consuming and prone to error. The present study uses both spatially variable hydraulic-head data and meteorological data to estimate groundwater input by LGD and lake water output through infiltration. The study sites are two shallow, groundwater-fed lakes without any surface inflows or outflows. Horizontally interpolated groundwater heads were combined with lake water levels to obtain vertical hydraulic gradients between the aquifer and the lake, which are separated by a thick layer of lake bed sediment which has a much lower hydraulic conductivity than the underlying aquifer. By fitting the hydraulic gradient to the results of a simple mass balance and considering the process of clogging, we were able to estimate the hydraulic conductivity of the lake bed sediments. We calculated groundwater inputs by LGD and lake water outputs by infiltration on an annual basis. Although our method requires several assumptions, the results are reasonable and provide useful information about the exchange between the aquifer and the lake, which can, for example, be used for the calculation of nutrient mass balances.

Lacustrine groundwater discharge (LGD) can play a major role in water and nutrient balances of lakes. Unfortunately, studies often neglect this input path due to methodological difficulties in the determination. In a previous study we described a method which allows the estimation of LGD and groundwater recharge using hydraulic head data and groundwater net balances based on meteorological data. The aim of this study is to compare these results with discharge rates estimated by inverse modelling of heat transport using temperature profiles measured in lake bed sediments. We were able to show a correlation between the fluxes obtained with the different methods, although the time scales of the methods differ substantially. As a consequence, we conclude that the use of hydraulic head data and meteorologically-based groundwater net balances to estimate LGD is limited to time scales similar to the calibration period.

No abstract