A comparative assessment of AWBM and SimHyd for forested watersheds

Abstract

The Australian Water Balance model (AWBM) and the SimHyd rainfall–runoff model are conceptual models widely used for simulating daily flows in Australia. To evaluate their ability to model non-stationary daily flows, to quantify the effect of land disturbance, and to assess their performance in catchments outside Australia, these two models were applied to two small watersheds, the Fernow watershed No. 6 in West Virginia, USA, for the period 1959–2009, and the River Rimbaud watershed in the French Alps for the period 1968–2006. Both watersheds have experienced well documented disturbances as a result of clearing and fire, respectively. The modelling protocol followed was adopted for a workshop on hydrology under change, held during the 2013 IAHS Assembly in Göteborg, Sweden, which was based on split-sample tests. On balance, the AWBM worked marginally better than SimHyd for these two catchments, and neither model worked satisfactorily for the Fernow watershed where forest clearing, application of herbicide and changes in species composition had occurred. There is little difference in terms of model performance between periods when land disturbances occurred and other periods with relatively stable conditions. Conceptual models are better equipped to simulate climate-driven variations in the observed streamflow (e.g. the River Rimbaud), and inadequate in reproducing streamflow variability as a result of complex forest management practices.