Climate Variability in the Southern Indian Ocean as Revealed by Self-Organizing Maps

Using a non-linear statistical analysis called “Self-Organizing Maps”, the interannual sea surface temperature (SST) variations in the southern Indian Ocean and their impacts on the southern African rainfall during austral summer are investigated. The SST anomalies during austral summer from 1951 to 2006 are classified into 9 types with different positions of positive and negative SST anomaly poles. To investigate the evolution of these SST anomaly poles, a heat budget analysis of mixed-layer temperatures is conducted by use of outputs from an ocean general circulation model. It reveals that the shortwave radiation anomaly mostly contributes to growth and decay of both positive and negative SST poles, and the contribution from latent heat flux anomaly is not so important in contrast to the results reported so far. This discrepancy is explained by the change in the contribution from the climatological heat flux term due to the interannual mixed-layer depth anomaly. For example, during austral summer when the mixed-layer is shallow, the positive contribution from the climatological shortwave radiation is amplified to enhance the warming trend in the positive SST anomaly pole, while the negative contribution from the climatological latent heat flux is amplified to suppress the warming trend. Also, we have found that large differences in the southern African rainfall anomalies exist owing to differences in moisture flux anomalies associated with the different positions of positive and negative SST poles. The present study suggests the importance of predicting the position of SST anomaly poles to enhancement of skills for predicting rainfall anomalies over the southern Africa.