Some Issues of Relevance to Operational Seasonal Forecasting(Invited Presentation)

Long term variations in the influence of the El Nino - Southern Oscillation events on Australian rainfall have been long recognised and documented by many authors. These secular changes are examined using a newly available monthly grided rainfall data set extending from 1900 to the present time. Correlations with the Southern Oscillation Index (SOI) over a moving 30 year window reveal large and rapid changes in the SOI - rainfall relationship over different parts of the continent in different seasons. These changes, which are shown to be due to a variety of causes, have major implications for seasonal rainfall forecasting, with widely different predictions possible, depending on the training period used to develop the forecast model.

In recent years the influence of other factors such as Indian Ocean Sea Surface Temperature anomalies (SSTa) on Australian rainfall have been documented. This influence is examined by stratifying the rainfall data using a classification based on the first two rotated Empirical Orthogonal Functions (EOFs) of the Indian and Pacific Ocean SSTa. The first rotated EOF is the large scale, mature phase of the El Nino, while the second rotated EOF describes variations in the subtropical eastern Indian Ocean. The stratified rainfall composites show that the Indian Ocean SSTa can significantly modify the effects of El Nino on Australian rainfall. In particular, there is a increased likelihood of experiencing above normal rainfall over southeast Australia during El Nino events in conjunction with warm SSTa in the Indian Ocean.