Preliminary analysis of trends in Australian flood data

Abstract

In recent years, the potential impacts of climate variability and change on the hydrologic regime have received a great deal of attention from researchers. Review of hydrological data recorded in different parts of the world has provided evidence of regime-like or quasiperiodic climate behaviour and of systematic trends in key climate variables due to climate change and/or climate variability. It has been established that a changing climate will have notable impacts on the rainfall runoff process, and thus hydrologic time series (e.g., flood data) can no longer be assumed to be stationary. A failure to take such change/variability into account can lead to underestimation/overestimation of the design flood estimate, which in turn will have important implications on the design and operation of water infrastructures. This paper presents preliminary results from a study aimed to identify the nature of time trends in flood data in the Australian continent with the final objective of assessing the impacts of climatic change on regional floods in Australia. This research is being carried out as a part of the on-going revision of Australian Rainfall and Runoff – the national guide of design flood estimation in Australia. For this study, 491 suitable stations with flood data in the range of 30 to 97 years have been selected across the Australian continent. Two trend tests are applied: Mann-Kendall test and Spearman’s Rho test to the data set. Preliminary trend analysis results show that about 30% of the selected stations show trends in annual maxima flood series data, with downward trends in the southern part of Australia and upward trends in the northern part. Further investigation is needed before any firm conclusion can be made about the trends in Australian flood data. Future work aims to address the influence of spatial correlation and autocorrelation on the ability to detect trend in annual maximum flood series data in Australia and assess the relationship between the observed trends in annual maximum flood data and other meteorological variables.