Progress on the enhancement of the Probabilistic Rational Method for Victoria in Australia

Design flood estimation for ungauged catchments is often required in design; commonly adopted methods for this include the index flood method, quantile regression techniques and various forms of the Rational Method. In Australian Rainfall and Runoff (ARR) 1987, the Probabilistic Rational Method (PRM) was recommended for general use in Victoria. The central component of the method is a runoff coefficient, which is assumed to vary smoothly over geographical space and across various average recurrence intervals. The PRM is widely used in practice due to its ease of application but also widely criticised due to the simplistic assumptions involved. Also, there has been a lack of independent testing of the PRM estimates and there is no clear indication of uncertainty associated with the method. This paper uses data from 133 catchments across Victoria to develop and test an enhanced PRM. A total of 20 test catchments were used for split-sample validation of the PRM. The streamflow data for these catchments were checked for trend, rating error and outliers. The selected catchments had streamflow record lengths in the range of 25 to 52 years. A log Pearson Type 3 — Bayesian at-site flood frequency analysis method was adopted in this study. From the independent testing, it was found that the enhanced PRM provides flood estimates of reasonable accuracy, with the median error in the range of 17 to 44% relative to at-site flood frequency estimates. The runoff coefficient values can be obtained from a new C10 contour map or alternatively from the developed prediction equation which requires four readily obtainable climatic and catchment characteristics variables. Generally, this prediction equation provides better results than those obtained from the C10 contour map