Including the influence of groundwater exchanges in a lumped rainfall-runoff model

The growing realisation that groundwater and surface water systems are components of connected hydrologic system has in recent years sparked the development of integrated surface-ground water models. In this paper, a version of the IHACRES rainfall-runoff model is presented, in which the CMD module for calculating effective rainfall is coupled to a streamflow-groundwater module, and applied to the Coxs Creek catchment, a variably gaining-losing stream system in Australia. The aim is to determine the capacity for the coupled model to capture the switching off-on behaviour evident in the observed flow record. Model performance can be improved in terms of event prediction, volume of baseflow and the percentile of flow cessation, through manipulation of CMD parameters, however, improvements in some performance criteria come at the expense of performance in others. An analysis of the input rainfall time-series, generated using a standard weighted Thiessen polygon approach, reveals mismatches between observed streamflow events and the occurrence of rainfall, which impose major limits on model performance. The challenge is to develop a simple lumped rainfall-runoff model that has the potential to improve system understanding and allow for meaningful exploration of alternate climate, groundwater extraction and land use change scenarios, given a situation of data poor catchments in many parts of Australia.