SYNOPSIS
Discriminant analysis is used to classify synoptic weather regimes at two contrasting locations in southern Africa for the purpose of wind power assessment. Four dominant weather categories at each location are identified and shown to be statistically distinct in terms of mean wind speed, mean wind power density and mean stability. Their differing characteristics and different frequencies of occurrence mean that the synoptic categories make vastly different contributions to total wind power at each of the two stations and, as such, represent means by which estimates of wind power can be extended by the use of numerical models, measurements can be organized into efficient samples and predictions of the surface wind fields can be developed.