An analysis of Australian seasonal rainfall anomalies: 1950–1987. II: Temporal variability and teleconnection patterns

Abstract A detailed examination of the variability of Australian district rainfall of seasonal time‐scales over the period 1950–1987 is described. In Part I, S‐ and T‐mode rotated principal component analysis (RPCA) was used to obtain a compressed representation of Australian district rainfall anomalies on seasonal time‐scales. In this paper, the temporal variability of these spatial patterns is examined, and their relationship to some large‐scale circulation indices, such as the Southern Oscillation (SO) and Pittock's L index. The S‐mode components show large intercomponent correlations as a result of the rotation of the principal components, although much of this is due to a common relationship to the SO. The T‐mode components are less strongly interrelated and only the first two show large correlations with the SO. The relationship with the SO and the L index are consistent with previous studies. Spectral analysis shows that most components do not have any dominant periodicity. Two distinct biennial tendencies are identified; in the north‐east associated with the SO, and in the south‐east associated with the L index. The large‐scale circulation anomalies associated with both the S‐ and T‐mode patterns are examined using the Comprehensive Ocean Atmosphere Data Set (COADS) and a network of rawindsonde stations over Australia and the South‐west Pacific. Three dominant teleconnection patterns are identified; the well‐known Southern Oscillation anomalies associated with the dominant T1 pattern of continental‐scale rainfall anomalies, a predominantly winter season Indonesian‐Indian Ocean pattern previously documented by Nicholls, and a spring/summer south Tasman Sea pattern. The evolution of these patterns is examined through composites, and the physical mechanisms for the initiation and maintenance of the anomaly patterns are discussed. In the latter two cases both the rainfall and SST anomalies appear to be driven by anomalies in the large‐scale mid‐latitude circulation.