We propose a novel method to evaluate regional palaeoclimate that can be used to alleviate the problems caused by the discontinuous nature of palaeoenvironmental data found in deserts. The technique involves processing satellite imagery and DEM's to map past rivers, catchments and evaluate the areas and volumes of palaeolakes. This information is used to determine the new Lake Evaluation Index (LEI) that allows a qualitative estimate of the amount of sediment received by lakes and how long-lived those lakes are. Lakes with considerable longevity and large sediment stores are selected for study. Validation is performed using image interpretation of remote sensing data, UltraGPR surveys and fieldwork. These techniques are also used to identify and study spring deposits and fluvial landforms that provide valuable palaeoclimate information. The method is applied to the Fezzan Basin in southern Libya focusing on the Wadi ash Shati and Wadi el-Agial catchments. Results indicate that the palaeohydrology is accurately mapped except within dune fields. We analysed the sedimentology of the key deposits identified by this methodology, developing a chronology using optically stimulated luminescence (OSL) and radiocarbon dating. We find evidence for relatively humid conditions during MIS 5c/d and e, as well as during the early to middle Holocene. Larger lakes and more extensive river systems were present during MIS 5 than are found during the Holocene, suggestive of greater humidity. The Holocene humid period started at ~11 ka and continued until ~5 ka being interrupted by abrupt periods of aridity at ~8.2 ka and ~6 ka that coincide with North Atlantic cooling. After each of these arid events the climate was less humid than previously, suggesting that they were superimposed upon 2 an overall drying trend. The termination of the Holocene humid period in the Sahara has received much scrutiny in recent years, and sediments of Palaeolake Shati provide a continuous record of this. We do not find evidence to support the hypothesis of either sudden or gradual aridification of the Sahara at ~5 ka, instead we find that that aridity started to develop at ~6.5 ka, whereupon the lake levels oscillated until finally drying-up by 5.3 ka. Most of the other lakes in the Fezzan also dried up at ~ 5ka. We suggest that thousands of years of aridification prior to 5 ka shrunk these lakes so that additional aridity at this time led to their final desiccation. Because lakes are prodigious dust sources this mechanism potentially explains the rapid rise in dust flux to the Atlantic at 5 ka, with this final drying being the culmination of longer term aridity, albeit overprinted with considerable climate variability.

No abstract