Assessing the value of Cl− and δ18O data in modelling the hydrological behaviour of a small upland catchment in northeast Scotland

Abstract

Model simulation of Cl− and δ18O in stream waters has been investigated as a means of improving interpretation of catchment-scale hydrological processes. The procedure has been evaluated for a small upland catchment which is one of the UK Environmental Change Network sites. Precipitation and stream samples have been analysed for hydrochemical determinands since the mid 1990s and, since November 2004, measurement of δ18O has also been undertaken. A conceptual hydrological model STREAM (STorage REsidence times And Mixing) was applied to the catchment to simulate the hydrology and responses of Cl− and δ18O. Results from model simulations confirmed that the catchment generally behaves as a well-mixed system. The feasibility of flow contributions from a deep groundwater source and infiltration excess runoff was examined, in addition to the apparently dominant shallow groundwater response. The ability to estimate mean residence times and draw strong conclusions about catchment processes was limited by the range of uncertainties in the experimental data and modelling. Integration of the tracer data in the model was found to be of value for probing model sensitivities and developing hypotheses that inform the design of further field experimentation. In this way, the modelling provides key feedback within a catchment learning framework.