Simplicity versus complexity in modelling groundwater recharge in Chalk catchments

Models of varying complexity are available to provide estimates of recharge in headwater Chalk catchments. Some measure of how estimates vary between different models can help guide the choice of model for a particular application. This paper compares recharge estimates derived from four models employing input data at varying spatial resolutions for a Chalk headwater catchment (River Pang, UK) over a fouryear period (1992-1995) that includes a range of climatic conditions. One model was validated against river flow data to provide a measure of their relative performance. Each model gave similar total recharge for the crucial winter recharge period when evaporation is low. However, the simple models produced relatively lower estimates of the summer and early autumn recharge due to the way in which processes governing recharge especially evaporation and infiltration are represented. The relative uniformity of land use, soil types and rainfall across headwater, drift-free Chalk catchments suggests that complex, distributed models offer limited benefits for recharge estimates at the catchment scale compared to simple models. Nonetheless, distributed models would be justified for studies where the pattern and amount of recharge need to be known in greater detail and to provide more reliable estimates of recharge during years with low rainfall.