Overcoming the challenges of using a rainfall–runoff model to estimate the impacts of groundwater extraction on low flows in an ephemeral stream

Simple modelling approaches such as a spatially lumped, rainfall–runoff model offer a number of advantages in the management of water resources including the relative ease with which groundwater and surface water accounts can be evaluated at the river-reach scale in data-poor areas. However, rainfall–runoff models are generally not well suited for use in ephemeral river systems because of their inability to simulate abrupt transitions from flow to no-flow periods and the highly non-linear rainfall–runoff relationships that exist in low yielding catchments. This paper discusses some of the challenges of using a rainfall–runoff model to assess the impacts of groundwater extraction on low flows within an ephemeral river system and demonstrates how these challenges were overcome during the development of the IHACRES_GW (Identification of Hydrographs And Component flows from Rainfall, Evaporation and Streamflow data – with Ground Water store) model. Details on the model algorithms, calibration, validation and objective function fits are provided. The performance of the IHACRES_GW model in Cox's Creek (Namoi Valley, Australia), and 13 additional areas investigated, suggests that this simple modelling approach may be of considerable utility for water accounting, especially when attempting to evaluate the impacts of groundwater extraction on low flows in similar systems.