The Tswaing Crater is around 40 km from Pretoria, South Africa. It was created by an asteroid impact some 200,000 years ago, which released roughly the energy of the Tunguska explosion of 1908. The crater's floor has a salt pan, where people have gone to gather salt since MSA times. The floor has been cored, with analyses of sediment salinity and pollen, giving a record of climate over the last 200,000 years. For example, a 2007 paper by Kristen and colleagues:
Sediments from Lake Tswaing (25°24'30'' S, 28°04'59'' E) document hydrological changes in southern Africa over the last 200 Ka. Using high-resolution XRF- scanning, basic geochemistry (TIC, TOC, TN), organic petrology and rock-eval pyrolysis, we identify intervals of decreased carbonate precipitation, increased detrital input, decreased salinity and decreased autochthonous (algal and bacterial) organic matter content that represent periods of less stable water column stratification and increased rainfall. Between 200 and 80 Ka BP, these intervals appear to be contemporaneous with local summer insolation maxima, indicating a strong influence of precessional variability (~23 Ka) on African subtropical climate. This influence weakens during the last glacial period (~80 to 10 Ka BP), when humid intervals at 73 to 68 Ka, 54 to 50 Ka, 37 to 35 Ka and 15 to 10 Ka BP are largely out of phase with insolation changes, and presumably reflect southward displacement of the ITCZ (Inter Tropical Convergence Zone) and/or changes in ocean circulation.
I'm pointing to this study because it is one that documents wetter periods during the span between the Howieson's Poort (roughly 60,000 years ago) and the Last Glacial Maximum (around 18,000 years ago). There are some who have claimed that this was a long span of aridity in southern Africa -- but more recent evidence makes it clear that the climate was not unimodal but fluctuated as in earlier and later time frames. Also, the climate was simply not arid, compared to "megadrought" periods documented in East Africa before 70,000 years ago.
Thanks to a reader, I've been reading an excellent 2008 paper by Peter Mitchell, which documents archaeological sites and paleoclimate data leading to the conclusion that habitation in southern Africa was not significantly interrupted during late MSA times. I'll refer to it more extensively later, but in the meantime I'm noting some recent work that Mitchell may not have had available when he was writing his article.
In a similar vein, I can point to an article by Louis Scott and colleagues (2008), which examined pollen records from Tswaing Crater as well as the Wonderkrater spring, and speleothem isotope evidence from Lobatse Cave, Botswana. Correlating the records from different sites in the same general area -- in this case, all from the savanna biome of southern Africa -- is very important. Distant climate records, such as the Greenland or Vostok ice cores, give some indication of global climate fluctuations, but it is not usually obvious how these fluctuations will affect specific regions of the world. One lake sediment core from southern Africa helps to show the local climatic fluctuations, but some may be highly localized, while others may reflect regional water and temperature variations. Hence, the correlations among many sites in a single region allow us to talk about climate fluctuations on a scale relevant to human populations.
Again, the region-wide picture in southern Africa between 60,000 and 20,000 years ago does not yield a picture of static, arid or cool conditions. The time period covers almost two full precessional cycles of insolation in southern Africa, and thus covers a wide range of local climate variation. To sum up, I'll cite Mitchell (2008:54), who relied on some earlier work from the Tswaing Crater record:
Peak annual precipitation may have reached 650–720 mm, with the late MIS 3 peak at the low end of this range. The minimum precipitation experienced was about 535 mm; today’s figure for comparison is 630 mm. Thus, although there were certainly periods when rainfall was reduced compared to the present , such reductions still exceeded by some margin the levels experienced by much of Limpopo Province or the highveld today, and for about a third of the time rainfall was actually higher than at present. Moreover, a generally cooler climate should have reduced evapotranspiration, and thus enhanced effective precipitation, more than these raw estimates suggest.
These sources are relevant for the savanna of the northern and eastern parts of South Africa. The region is ecologically diverse, and Mitchell considers different parts of the region in turn.
Kristen I, Fuhrmann A, Thorpe J, Röhl U, Wilkes H, Oberhänsli H. 2007. Hydrological changes in southern Africa over the last 200 Ka as recorded in lake sediments from the Tswaing impact crater. S Afr J Geol 110:311-326. doi:10.2113/gssajg.110.2-3.311
Mitchell P. 2008. Developing the archaeology of marine isotope stage 3. S Afr Archaeol Soc Goodwin Ser 10:52-65.
Scott L, Holmgren K, Partridge TC. 2008. Reconciliation of vegetation and climatic interpretations of pollen profiles and other regional records from the last 60 thousand years in the Savanna Biome of Southern Africa. Palaeogeography, Palaeoclimatology, Palaeoecology 257:198-206. doi:10.1016/j.palaeo.2007.10.018