Canonical correlation analysis (CCA) is used to downscale large-scale circu
lation forecasts by the Centre for Ocean-Land-Atmosphere studies (COLA) T30
general circulation model (GCM) statistically to regional rainfall in Sout
h Africa. Monthly GCM ensemble forecasts available from 1979 to 1995 have b
een generated using NCEP reanalysis data as initial input and globally obse
rved sea-surface temperature (SST) data at the lower boundary. Altogether,
51 30-day cases of CCM simulations, spanning 17 years, within the target se
ason of December-February (DJF), are produced. This period is very importan
t for agriculture and maize, in particular. A model output statistics (MOS)
procedure is used to downscale GCM forecast sea-level pressure and 500 hPa
height fields to regional rainfall for 30-day periods over South Africa. T
he CCA model is trained on the first 31 cases (up to February 1989) and for
ecasts are subsequently made for the remaining 20 cases. These retro-active
real-time forecasts have a high potential (correlations > 0.5) over most o
f the interior of South Africa and, furthermore, the prediction of extreme
events seems feasible. CCA diagnostics of the GCM-output against rainfall r
eveal that favourable rainfall over most of the interior is associated with
low pressure systems at the surface over the west coast, with an associate
d ridging high. This is supported by other observational studies. Copyright
(C) 2000 Royal Meteorological Society.