Flood frequency analysis : a case study for the Brisbane River catchment

Abstract

Flood is the most common natural hazards around the globe that has notable negative effects on humans and environment. One of the examples is Queensland 2010-2011 flood, which is considered as one of the severest floods in recent history of Australia that claimed 31 human lives and caused direct damage costing over $5 billion. To reduce the flood damage, it is vital to understand properly the causes of major floods, their magnitudes and frequencies. Estimation of the magnitude of possible future floods (also called design floods) is an important task in hydrology. Most of the hydraulic structures and flood management tasks require an accurate estimation of design floods. For this reason, estimation of design flood is still an area of great interest in flood hydrology and is being researched worldwide. Frequent devastating floods in Australia have drawn attention at the state and national levels for more accurate flood estimation with reduced uncertainty. Many design floods estimation methods are being practiced around the world. This study focuses on the widely used design flood estimate techniques called "flood frequency analysis (FFA)". The main objective of FFA is to find probability distribution model that best fits the measured flood data series at a given site. Although Australian Rainfall and Runoff ARR (Australian Rainfall and Runoff), 1987 recommended Log Pearson type III probability distribution to use for FFA in Australia, in ARR 2019, no specific probability distribution is recommended. There has been limited guideline in Australia to select probability distribution models for flood frequency analysis. Also, many users have limited understanding on the uncertainties involved in design flood estimates based on a given probability distribution. This study is devoted to fill this research gap and examines the selection of the most appropriate probability distributions and associated uncertainty in FFA. This study focuses on the Brisbane River catchment of Queensland, one of the worst flood-prone areas in Australia. In this research, a total of 26 streamflow gauging stations are selected from the Brisbane River catchment, with the lengths of recorded annual maximum flood (AMF) data series in the range of 20 years to 91 years.

Date of Award	2020
Original language	English