Numerical Modelling Of Severe Thunderstorms Over The South African Highveld With The Aid Of The Clark Cloud Model

Abstract

Precipitation research, with a view to rainfall augmentation, especially in a drought-stricken country like South Africa, is an attractive option as a source of good quality water. Numerical models can be useful tools in these programmes to help investigate the dynamical and microphysical processes in clouds. They can also be of great benefit to better understand the outcome of cloud seeding and to assist in the formation of a seeding hypothesis. This dissertation serves as a first step in establishing if the Clark model could predict mesoscale conditions in South Africa accurately enough, in order to be used as a tool in the precipitation research programme. The Clark model was used to simulate the mesoscale conditions of two selected days over the Bethlehem area. On both case days severe storms were responsible for extensive hail damage over the area. It was assumed that if the model could accurately predict severe conditions, it would also be able to predict general mesoscale conditions (this assumption was proved in the dissertation). The model also highlighted the fact that a single initializing sounding would not always be sufficient to accurately simulate severe convective activity. The model was also used to study the effect that surface heating and topography have on the development of convective activity.