Phillips, Professor David1, Matchan, Dr Erin1
1School of Earth Sciences, The University of Melbourne, Parkville, d9cec488-596f-4927-97dc-c72b3e7f8ea7
Archaeological expeditions by the National Museum of Kenya to east Lake Turkana (formerly Lake Rudolf) in 1968 and 1969 led to the discovery of remarkable stone artefacts and hominin fossils, associated with volcanic tuffs, including the famous KBS (Kay Behrensmeyer Site) tuff.
Initial attempts to date the KBS tuff proved unsuccessful due to fluvial deposition of most tuffs and contamination by older material. Early 40Ar/39Ar dating of pumice feldspar grains from the KBS tuff yielded a reported age of 2.61 ± 0.26 Ma, which was soon disputed on the basis of faunal correlations, thereby precipitating a controversy that played out in Nature publications for the next decade. In 1975, the Berkeley geochronology laboratory published K-Ar ages of 1.60 and 1.8 Ma for two KBS localities, with the former age later attributed to laboratory error. The Cambridge geochronology laboratory then reported a range of 40Ar/39Ar ages (0.52 – 2.61 Ma) and revised the age of the KBS tuff to 2.42 ± 0.02 Ma, in accord with preliminary zircon fission track ages. Subsequent geochemical correlations with the H2 (=KBS) tuff in Ethiopia, dated at ca. 1.8 Ma by the K-Ar method, heightened the controversy. Later K-Ar and 40Ar/39Ar dating analyses by Ian McDougall at the Australian National University largely resolved the debate, with reported ages of 1.89 ± 0.01 Ma and 1.88 ± 0.02 Ma, coincident with a revised fission track age of 1.87 ± 0.04 Ma. More recent analyses of single feldspar crystals from KBS pumice clasts produced a weighted mean age of 1868 ± 14 ka – but with significant scatter, suggesting the presence of inherited grains or extraneous argon.
New ultra-precise 40Ar/39Ar data obtained for single feldspar pumice crystals from several tuff localities across the Omo-Turkana Basin, including the KBS tuff, show variably complex age distribution patterns even within single pumice clasts. Based on co-irradiated A1T and FCT sanidine aliquots and a Bayesian statistical analysis approach, we calculate astronomically calibrated ages for several tuff horizons at precision levels approaching <<0.1%. The KBS and correlated H2 tuff give an astronomically calibrated, weighted mean age of 1879.2 ± 1.3 ka (0.069% 95%CI). The stratigraphically younger Malbe (=H4) tuff, which was originally misidentified as the KBS tuff, gives a Bayesian eruption age of 1837.75 ± 0.86 (0.047%; 95%) ka. The These results enable effective stratigraphic correlations across the Basin, and reveal paleoclimate and paleo-environmental variability at millennial timescale resolution.
This study is dedicated to the memory of Ian McDougall and Frank Brown, who worked tirelessly to unravel the magmatic and geological history of the Omo-Turkana Basin.
Professor David Phillips is Head of the School of Earth Sciences and Director of the 40Ar/39Ar Laboratory at the University of Melbourne. He is internationally recognised for his research on ultra-precise 40Ar/39Ar dating methods and their application volcanic rocks, including tuffs related to hominin localities in the Turkana Basin, Kenya.