Last millennium climate reconstruction in Ethiopia using multiple stalagmite parameters

The Ethiopian highlands are a classic example of a rain-sensitive region where future forecasts are hampered by inadequate understanding of historical patterns, their wider associations, and causes. Rainfall and temperature records are relatively short and of poor quality. The few long records demonstrate that after El Niño years, the spring rains are heavy and the main summer rains are reduced, whereas in normal / La Nina years, spring rains are variable / reduced and the summer rains are stable. However this relationship is too weak to enable drought forecasts and the calibration time period is too short. In addition, the spring rains also fail and lead to famine; their cause is more complicated and spring rainfall failures are, to date, not predictable. This summer, unpredicted excessive summer rains caused flooding over most of northern hemisphere tropical Africa, including Ethiopia. An enhanced seasonal precipitation range is predicted under climate model scenarios of future climate change Long, high resolution climate records are required to investigate the nature of rainfall variability, the frequency of failure of either rain, as well as investigate the presence of longer term periodicity in climate that cannot be detected through short instrumental series. A particularly powerful approach will be to make use of well-dated proxy material to derive independent tests of past climate variability. Recent rapid advances in the understanding of the climatic meaning of parameters derived from calcareous speleothems (cave precipitates), coupled with our pilot work in Ethiopia, indicates that speleothem records are the best prospect for tackling this urgent problem in Ethiopia, as to date other potential proxies have failed to yield high resolution climate reconstructions. Here we propose to derive palaeoclimatic parameters such as mean annual precipitation, seasonality and annual moisture excess through the multiparameter analysis of speleothems that have grown over the last millenium. Uncertainty on each of the individual parameters will be derived from the noise in analytical series and by repeating analyses of different proxies on different samples from the same geographic location. We therefore aim to reconstruct climate series as quantitative parameters that can be modelled by, and therefore used to validate, climate models.