On Australian heat waves: Time series analysis of extreme temperature events in Australia, 1950-2005

The occurrences of extreme weather, particularly heat-waves can often result in significant costs in terms of human health, damage to infrastructure, increased stress of natural ecosystems and reduced economic activity. In Australia, temperature extremes, producing intermittent heat-waves silently kill almost as many people as do any other natural disasters (Granger and Berechree 2005). \nWhile there appears to be no official definition, a heat-wave may be considered as a prolonged period of excessively hot conditions that usually lasts more than three consecutive days. This definition, which includes the combined effect of high temperatures and humidity is based on Stedman’s (1979) approach and allows us to compute the relevant heat wave index to assess the severity of heat-related events. \n \nThis paper presents a statistical analysis of daily maximum temperatures and humidity records for selected Australian cities and rural stations, with a focus on the trends of Australian heat waves since 1950. The two datasets were used to extract the time series of “maximum apparent body temperatures” (also referred to as the yearly-worst heat wave indices) in order to quantify the trends, spatial distribution and statistical significance over the continent. The Mann-Kendall’s (MK) nonparametric \ntest and Sen’s slope estimation techniques were used to quantify the overall statistical significance of the results.