Numerical simulations of a severe rainfall event over the Eastern Cape coast of South Africa: sensitivity to sea surface temperature and topography
Article first published online: 6 APR 2006
Volume 58, Issue 3, pages 355–367, May 2006
How to Cite
SINGLETON, A. T. and REASON, C. J. C. (2006), Numerical simulations of a severe rainfall event over the Eastern Cape coast of South Africa: sensitivity to sea surface temperature and topography. Tellus A, 58: 355–367. doi: 10.1111/j.1600-0870.2006.00180.x
- Issue published online: 6 APR 2006
- Article first published online: 6 APR 2006
- (Manuscript received 2 March 2005; in final form 9 Jaunary 2006)
The intense precipitation event over East London, South Africa, on 15–16 August 2002 is investigated using a mesoscale model (MM5) and its sensitivity to sea surface temperature (SST) and topography tested. Over 300 mm of rain fell in 24 h over East London as compared to the August average of 78 mm. The intense precipitation resulted from a cut-off low system aloft and the formation of a low-level jet (LLJ) near the surface.
The MM5 control simulation indicated that the intense precipitation was due to the ascent of moist air at the leading edge of the LLJ, which advected warm moist air at low levels onto the coastal mountains and was enhanced by the formation of a meso-beta scale low-pressure system just off the coast. The ascent at the leading edge of the LLJ formed the ascending arm of a convective circulation cell.
Sensitivity experiments with MM5 indicated that the warm core of the Agulhas Current contributed to the formation of the meso-beta scale low, while the mesoscale structure of the SST field was important in determining its location and that of the LLJ, and therefore the precipitation in the model. Most of the moisture originated from the Agulhas Current region. The South African topography was found to influence the location of the LLJ via trapping a near-surface warm trough on the eastern side of the Drakensberg mountains. We suggest that deficiencies in the control experiment may have been due to inaccuracies in using a weekly mean SST field to force the model at the lower boundary.