To isolate the most important time scales in the SST-rainfall relationship, a bandpass filtering of the Hulme and New rainfall and UKMO GISST2.2 SST data sets are performed. The particular interannual, quasi-decadal and interdecadal bands chosen for this purpose are motivated by a joint EOF study by other workers of global modes of low frequency variability using UKMO GISST3.0 and UKMO/CSIRO GMSLP3 data sets. Correlations between these bandpass filtered SST and rainfall data sets indicate a statistically significant relationship between South West Indian Ocean SST and South African summer rainfall on both interannual and decadal scales.
The results of this analysis are used to motivate AGCM experiments using an idealisation of the observed SST anomalies. The AGCM response to these anomalies includes changes in regional atmospheric circulation and moisture transport over southern Africa that appear consistent with previously identified wet and dry season circulation patterns; hence providing further support for the observational evidence of a link between rainfall and SST in the region.
![](images/flag.png)
![](images/blue_ribbon.png)