Impact of data uncertainty on the spatio-temporal distribution of Australian cyclones, 1906-2006

Tropical cyclones (known elsewhere as hurricanes or typhoons) regularly track through Australia’s northern coastal regions (Figure 1). Data documenting the basic characteristics of cyclones in the Australian region is available from the Bureau of Meteorology (BOM) as early as 1906. This data set (freely available in ASCII format at http://www.bom.gov.au) records some 41 attributes (most of which are not complete for the entire time series) for 22,645 individual cyclone eye positions. Because cyclones can cause considerable impact to both marine and terrestrial environments, data sets such as this are widely used to reconstruct past cyclone conditions (for example, see Berz et al 2001, Peduzzi et al 2005, Puotinen 2007) to enable comparisons with what may happen in the future under global climate change (i.e., Webster et al. 2005, Walsh et al. 2004). Because cyclones tend to vary on long time scales (centuries or more), the accuracy of this type of modeling is sensitive to the length of the time series on which the analysis is based (Nott 2003). Thus, long time series are of particular value.