Towards a new regional flood estimation method for the Northern Territory

![CDATA[This paper presents the development of a regression based regional flood estimation method for the northern part of the Northern Territory as a part of the ongoing research in relation to the upgrade of regional flood estimation methods in Australian Rainfall and Runoff. A total of 51 catchments are used in this study. Using the ordinary least squares regression, predictions equations are developed for the average recurrence intervals (ARIs) of 2, 5, 10, 20, 50 and 100 years. These prediction equations contain two predictor variables; catchment area and mean annual rainfall, which are relatively easy to obtain. The developed prediction equations satisfy the model assumptions quite well. An independent test shows that the developed prediction equations can provide quite accurate flood quantile estimates with median relative error values in the range of 32% to 45% as compared to at-site flood frequency estimates. Further study will involve application of the generalised least squares regression technique to the data set and regionalization of the parameters of the log Pearson Type 3 distribution. The findings from this and other on-going studies will form the basis of recommendation of new regional flood estimation methods in the Australian Rainfall and Runoff for the Northern Territory.]]