Flood Estimation in Northern Australian Catchments using Monte Carlo Simulation Technique

A number of recent studies have shown that the currently used Design Event Approach to rainfall-based design flood estimation offers only limited scope for further improving the accuracy of design flood estimates. These limitations can be largely overcome by a new Monte Carlo simulation technique, where a large number of runoff events are generated by drawing random values from respective probability distributions of model inputs/parameters. These stochastic runoff events are then routed through a calibrated runoff routing model, and the simulated hydrographs (in the order of thousands) used to construct a derived flood frequency curve. The recent application of this technique to three small Victorian catchments had shown promising results. This paper assesses the performance of the Monte Carlo simulation technique under the different hydrologic conditions that exist in Northern Australia. Based on the application to three small Queensland catchments, it has been found that, compared to Victoria, there are larger numbers of rainfall events with longer durations in Northern Australia. Also, initial losses are found to be much higher in Northern Australian catchments. The results show that the Monte Carlo simulation technique is readily adaptable for application to Northern Australian catchments.