Climate variability and predictability in tropical Southern Africa with a focus on dry spells over Malawi

Abstract
The climate variability and predictability in tropical southern Africa at inter-annual time scale is studied. Dry spells are identified using Malawi daily rainfall and circulation patterns are investigated at intraseasonal time scale. The study first examines the temporal and spatial variations of Malawi summer rainfall and relationships with global/regional environmental variables are established. Malawi summerrainfall shows that extreme weather has become more frequent in the last three decades. Three oscillations are identified in Malawi summer rainfall; these are the 2.4 and 3.8 year cycles followed by the 11.1 year cycle. These cycles may be associated with the Quasi-Biennial Oscillation (QBO), El Nino/Southern Oscillation (ENSO) and solar cycle respectively. The summer rainfall indicates strong correlation with Nino3 sea surface temperature and Indian Ocean outgoing longwave radiation (OLR), three to six months prior to the rainy season. Malawi pentad summer rainfall at intraseasonal time scale shows similarity in distribution with other regional pentad summer rainfall, (e.g., onset and cessation dates). Results from spectral analysis of daily rainfall indicate two major cycles, namely the 10-25 and the 30- 60 day cycles. Results from composite analysis, using the NCEP reanalysis data set, reveal distinct circulation patterns prior to occurrence of dry summers over Malawi. The circulation patterns prior to dry summer are dominated by westerly flow which changes latitude, causing subsidence. Below (above) normal sea surface temperatures are observed across the east Atlantic(central Indian Ocean), a pattern typical of El Nino conditions. Pronounced below normal geopotential heights occur during dry summers to the south of Africa and east of Madagascar with corresponding south westerly wind anomalies. Results from case studies of dry spells at intraseasonal time scale indicate the prominence of deep low pressure cell east of Madagascar and the absence of a well defined Inter-tropical Convergence Zone (ITCZ) over southern Africa. Dry spells are dominated by high pressure cell over Botswana in conjunction with divergent southerlies and subsidence over southern Africa. Results from composites reveal possible predictors for Malawi summer rainfall: zonal winds, geopotential heights east of Madagascar, SST in the central Indian Ocean and in the south east Atlantic Ocean, and low-level velocity potential over northern Madagascar. Results from multivariate regression analysis show that lowest predictability is found in early Malawi summer rainfall (NDJ) while highest predictability is found with FMA rainfall. The highest predictability for November-April (NA) Malawi summer rainfall is associated with sea surface temperature in the southeast Atlantic Ocean and Indian Ocean, sea level pressure over Indian Ocean and QBO in JAS months. This study has contributed to the understanding of summer rainfall in tropical southern Africa. The knowledge gained can be used by decision makers, farming community, water resource managers for planning and operational purposes. Further statistical forecast models could be developed from precursors (predictors) identified in the study to assist in mitigating the negative effects of climate variability.
Description
Thesis submitted to the Faculty of Science for the degree of Master of Science in the Department of Geography and Environmental Studies at the University of Zululand, South Africa, 1999.
Keywords
Climate variability, Climate predictability
Citation