Southern African palaeoclimates and variability : the story from stalagmites, pollen and coral.

Thesis (M.Sc.)-University of KwaZulu-Natal, Durban, 2005.

Compared to extensive study in the northern hemisphere, very little is known of southern African palaeoclimates. This study aimed to extend understanding of the nature of and controls on southern African palaeoclimates of the last 40 000 years. Through a study of the approximately 20 000 year long Makapansgat and Wonderkrater palaeoclimatic records, and an extensive literature review of southern African palaeoclimatic studies, a number of common rainfall and temperature fluctuations were detected across the summer rainfall region. Based on these trends, general models of rainfall and temperature changes over time were developed for the region. The analysis of a coral core, derived from a Porites lutea head from Sodwana Bay, covering the last 116 years, indicated higher frequency climatic fluctuations over the last century. Climatic variability on the long- and short-term could then be related to known atmospheric processes through application of the Tyson (1986) model for southern Africa atmospheric circulation. North-south shifts in mean circulation dominate climatic variability in the region but there are also regular disturbances to this mean, such as in the form of the EI Nino - Southern Oscillation. The fluctuations seen in present and palaeoclimatic records are the result of a complex interaction between internal and external mechanisms of climate change. Wavelet analyses of recorded and proxy climatic datasets highlighted the cycles which dominate southern African climatic variability on timescales from years to millennia. The causes of these cycles were then assessed in the context of established solar, atmospheric and oceanic models. Wavelet analyses also provided an indication of frequency changes over time and were therefore useful for detecting climate change. An analysis of proxy and recorded climatic datasets for southern African rainfall over the last 100 years indicated a frequency modulation of the 18 year rainfall cycle, which was first described by Tyson (1971). This variation may be related to anthropogenic climate change. It became apparent from this study that there is a need for increased scientific interest in the palaeoclimatic trends of the region. The number of continuous, high-resolution datasets needs to be increased to allow for comparison and confirmation of various trends with records from sites across the globe. An understanding of the nature of regional and global teleconnections is essential before reliable climate change models can be established. There is also a need for further understanding of short-term southern African climate variability on inter-annual timescales.It is only once we have an understanding of the natural climatic variability of the region ,and its inherent cyclicity,that we can begin to distinguish the impact of anthropogenic activities on climate.