Detection of trends in extreme rainfall events in NSW, Australia

Climate change and variability affect rainfall frequency and magnitude. It has been reported that south-east Australia has been experiencing declined mean annual rainfall due to climate change. Trend detection in rainfall time series is one of the most important research areas in hydrology and climatology to investigate the impacts of climate change on rainfalls. This study focuses on detection of trends of sub-hourly, sub-daily and daily extreme rainfall events in New South Wales (NSW), Australia. It uses data from 20 pluviograph stations in NSW covering rainfall data from 1976 to 2012. Two non-parametric tests, Mann-Kendall (MK) and Spearman Rho (SR) tests are applied to detect trends at 1%, 5% and 10% significance levels for fifteen different storm durations ranging from 6- minute to 72 hours. Pre-whitening (PW), Trend-Free Pre-whitening (TFPW) and the Variance Correction (VC) approaches are adopted to account for the impacts of serial correlation in the annual maximum rainfall series on results of the trend analysis. It has been found that for short durations, the statistically significant upward trends are evident in greater number of stations than the statistically significant downward trends. Depending on the results of both the MK and SR tests, it is found that the MK and SR tests provide quite different results for some storm durations considered in this study. Use of PW, TFPW and VC approaches, which are adopted to remove the impact of serial correlation on trend results, the percentage of stations having significant upward trend have been slightly reduced. The findings of this research will help to assess the impacts of climate change on design rainfall intensity-frequency-duration (IFD) data in NSW.