Lake-based climate reconstruction in Africa: progress and challenges

Lake sediments are and will continue to be the principal source of information on the climate history of tropical Africa. However, unequivocal interpretation of the various sedimentological, biological, and geochemical climate-proxy data extracted from lake sediments with respect to past variations in temperature, rainfall, and wind is an extremely complex and challenging exercise. Outstanding problems are: (1) the inherent conflict between a lake's sensitivity to climate change (its ability to respond to and record relatively modest, short-lived climatic anomalies) and its persistence as an archive of climate change (the probability that it survived the most arid events without desiccation or erosion, allowing it to preserve a continuous record of climate history); (2) the scarcity of annually laminated sediment records, which in other regions can provide superior chronological precision to lake-based climate reconstructions; (3) lack of a quantitative (sometimes even qualitative) mechanistic understanding of the chain of cause and effect linking sedimentary climate-proxy indicators to particular climatic variables; and (4) lack of a proxy indicator for past temperature changes unaffected by simultaneous changes in moisture balance. Clearly, a climate-proxy record with high stratigraphic resolution does not represent a high-resolution record of past climate change without demonstration that the sedimentary archive is continuous and undisturbed; that the lake system responds to climate variability at the appropriate time scale; and that any threshold effects in the relationship between the proxy indicator and climate are accounted for. Calibration and validation of climate-proxy indicators is tantamount to establishing accurate reconstructions, but in Africa historical validation of proxy indicators is handicapped by the scarcity of long-term lake-monitoring data. The reliability of lake-based climate reconstructions is enhanced when inferences derived from several proxy indicators (sedimentological, biological, or geochemical), that each have an independent mechanistic link to climate, show a high level of coherence. Given the scarcity of annually-resolved sediment records in tropical Africa, we may have to accept the limitations of ²¹⁰Pb- and ¹⁴C-based chronologies when evaluating the synchrony of reconstructed climate events between sites and regions; however, careful site selection and detailed lithostratigraphic analyses can go a long way to optimise depth-age models and reduce uncertainty in the timing of past climate changes.