Regional flood estimation in Australia : an overview of the study for the upgrade of 'Australian Rainfall and Runoff'

Regional flood frequency analysis (RFFA) methods are used to estimate design floods in ungauged catchments. In Australia, the RFFA methods have not been upgraded since 1987. The National Committee on Water Engineering has commenced a major upgrade of the Australia Rainfall and Runoff (ARR). In the context of this upgrade, Project 5 'Regional Flood Methods' focuses on collation of a national database, development and testing of various RFFA methods, which will form the basis of recommendations on new RFFA methods in the revised edition of ARR. This paper presents an overview of the progress with Project 5 in relation to the ARR revision. It covers data preparation, testing of various RFFA methods, adoption of final methods and the development of an application tool. In the development of the final method, a total of 676 gauged catchments have been used to create six regions covering Australia. A Bayesian GLS regression technique has been adopted to develop prediction equations. A regionalised LP3 distribution is recommended to derive flood estimates for ungauged catchments in the range of 2 to 100 years ARI. For all the regions, the proposed prediction equations used only two predictor variables, catchment area and design rainfall intensity. An application tool is under preparation which will automate the application of the new RFFA method, referred to as 'ARR RFFA 2012'. The user will be required to input latitude, longitude, catchment area and design rainfall intensity to obtain design flood quantiles and associated uncertainty estimates. It should be noted that the developed ARR RFFA 2012 method will need to be re-calibrated with the soon-to-be-released revised IFD data from ARR Project 1 before it can be applied by the industry.