A model for the identification of tropical weather systems over South Africa

Abstract

South Africa forms the southern end of Africa with its northern boundary at approximately 22°S and the southern-most point, Cape Agulhas, at approximately 35°S. During most of the year atmospheric circulation over South Africa, especially the central and southern regions, is dominated by extra tropical weather systems such as cut-off lows, cold fronts and the ridging Atlantic Ocean High. Weather forecasters in South Africa are trained on and experienced in forecasting rainfall from these systems. During late summer weather systems from the tropics invade the northern regions of South Africa. Because tropical weather systems only appear during three or four months of the year weather forecasters often lack the experience to identify them timeously. Invading tropical weather systems are often associated with heavy rainfall and flooding (February 1988, February 1996, February 2000). It is very important that forecasters identify these tropical systems and apply appropriate forecasting techniques. Prior to this research no objective system to classify tropical systems and to pinpoint areas of significant to heavy rainfall was employed locally. This research resulted in the model for the identification of tropical weather systems (MITS). MITS has five components based on the atmospheric dynamics important for the development of convective rainfall from tropical systems. The application of MITS is detailed in a case study of heavy rainfall from a tropical system over the country on 8 February 2000. MITS identified the tropical nature of the system and successfully located areas of convective rainfall. MITS is available on the software system PCGRIDDS currently in operational use in South African forecasting offices. It operates on the prognostic fields generated by the South African Weather Service limited-area model Eta. Trial runs of MITS in the Central Forecasting Office of the South African Weather Service (SAWS) provided a positive response from weather forecasters and a full operational implementation is imminent.


WaterSA Vol.28(3) 2002: 249-258