Trend analysis of rainfall losses using an event-based hydrological model in eastern NSW

Abstract

Climate change is likely to impact on the magnitude and frequency of extreme flow events. Whilst flood data have traditionally been assumed to remain stationary, this may no longer be the case due to the effects of climate change. A number of studies have demonstrated trends and variability in annual maximum flood data across Australia. These findings have been supported by Australian climate change studies, which predict increases in extreme rainfalls, increasing evaporation and changes to soil moisture conditions. However, what it is more uncertain is the magnitude of these changes, the specific changes at a regional scale and the impact on catchment characteristics at the local scale and how this affects catchment response when a hydrologic model is to be developed. This study investigates trends in loss parameters using an event-based rainfall-runoff model, for four catchments across eastern New South Wales. These catchments include the Orara River, Ourimbah Creek, Currambene Creek and Pambula River. The analysis explores three loss models, the initial loss-continuing loss (IL-CL) model, initial loss-proportional loss (IL-PL) model and Soil Water balance MODel (SWMOD). Trends were investigated for the loss parameters of all three loss models, including the initial loss (IL), continuing loss (CL), proportional loss (PL) and initial moisture (IM, from SWMOD). The relationship between trends in loss parameters and trends in the annual maximum flow series, antecedent precipitation index, annual runoff totals, annual rainfall totals, number of rainy days and inter-event duration are also examined. In this study, four tests were adopted to test for trends, namely the Mann-Kendall, linear regression, CUSUM and cumulative deviation tests. Results from the tests of change and visual analyses show that for two of the four catchments there is strong evidence of an upward trend in the IL data, strong evidence of a downward trend in the IM data and strong evidence of a downward trend in the API data (28 days). The changes in the initial condition for rainfall-runoff modelling (IL and IM) correspond well to the trends in the API, as surrogate for the antecedent moisture content. Only one of the catchments saw a change in the continuing loss measures (CL and PL), however, this is also the only catchment that shows evidence of trends in the annual flood series.