Database underpinning the RFFE model 2015 in the new Australian rainfall and runoff

The new regional flood frequency estimation (RFFE) model for Australia, known as 'RFFE Model 2015' is based on an extensive database of gauged catchments. This paper presents important characteristics of the database so that the users of the RFFE Model are aware of the strengths and limitations of the database and how these could impact the outcomes of the RFFE Model. The RFFE Model data set consists of 798 gauged catchments from the humid coastal areas and 55 catchments from the arid/semi-arid areas. Humid coastal areas include most of the coastal belts of Australia where there are a good number of recorded streamflow stations. The arid/semi-arid areas include Pilbara and interior Australia. An upper limit of catchment size of 1,000 km2 is adopted; however, in Tasmania and in the Northern Territory, a few larger catchments are included as the numbers of catchments with areas less than 1,000 km2 are too small. The cut-off streamflow record length is selected as 19 years to maximise the number of eligible stations on the consideration that a higher cut-off would reduce that number. The annual maximum flood data sets for the selected catchments are collated based on a stringent data screening procedure e.g. gaps in the annual maximum flood series are in-filled as far as could be justified, outliers are detected using the multiple Grubbs-Beck test and error associated with rating curve extrapolation is investigated. The at-site flood frequency analyses are conducted using the FLIKE software. The potentially influential low flows are identified using multiple Grubbs-Beck test and are censored in the flood frequency analysis. A Bayesian parameter estimation procedure with LP3 distribution is used to estimate flood quantiles for each of the gauged site for annual exceedance probabilities of 50% to 1%. A total of five catchment predictor variables are included in the RFFE Model. The database prepared for ARR Project 5 is a valuable resource in Australian hydrology, which is expected to be progressively updated.