Tropical cyclones (TCs) pose a recurring threat to the Australian region (90-160° E) each year. The Australian TC season extends from November to April, with a peak in TC activity observed typically during the months of February and March. The interannual variability of TC activity in the region is strongly associated with the El Nino Southern Oscillation (ENSO) phenomenon, and on average there are less TCs during the warm phase of ENSO (El Nino) and more TCs during the cool phase (La Nina). Previous studies have employed a range of ENSO and also non-ENSO-related individual parameters to establish a baseline relationship between the particular parameter and annual TC activity. The ENSO-related parameters, for example, have included Darwin sea level pressure and the Southern Oscillation Index (SOI), and both have been used extensively. However, to our knowledge, no study has yet investigated the best set of parameters for such correlations.

The present study explores the predictability of Australian TC activity using the following five ENSO-related parameters: (i) Nino 3.4 SST, (ii) Nino 4 SST, (iii) SOI, (iv) Darwin pressure and (v) the leading principal component (PC) of Pacific SSTs; by systematically correlating those parameters with TC frequency over a 35 year period, for a range of lead and lag times. The parameters are also correlated with each other, to assess their degree of interdependency. Preliminary results suggest that the Nino 3.4 and Nino 4 SSTs are very strong candidates for the prediction of TC frequency in the region, even several months prior to the onset of the TC season. This possibly is a consequence of the persistence of tropical SST anomalies in the Pacific, once established, compared to the relatively large fluctuations in atmospheric pressure. The geographic location of TC genesis appears to be only weakly related to ENSO, contrary to what previous studies appear to suggest.