COMPARISON OF AGES FROM U-SERIES ON PEDOGENIC CARBONATE AND 10BE DEPTH PROFILES FOR TWO LATE PLEISTOCENE ALLUVIAL FANS IN SOUTHERN CALIFORNIA

Accurate and precise dating of alluvium in arid environments is an enduring challenge, but is essential for many neotectonic and geomorphic studies. We combine U-series on pedogenic carbonate and 10Be depth profile modeling, and apply them to two, late Pleistocene alluvial fans in southern California. U-series analyses of pedogenic carbonate coating gravel-clasts yield reliable minimum ages for both fans, as indicated by good agreement in ages and back-calculated initial 234U/238U among multiple samples. In the Imperial Valley, a surface with remnant bar and swale topography has a weighted mean U-series age of 41.4 ± 1.6 ka (n=7, probability =0.11). A surface in the Anza Borrego Desert that preserves a Pleistocene relict A soil horizon (indicating very little erosion) has a weighted mean U-series age of 33.3 ± 1.4 ka (n=4, probability =0.8). To analyze the 10Be depth profiles we use a model (Braucher et al., Quat. Geochron., 2009) that determines the best solution for the age – erosion rate – inheritance triplet. Our 2-meter-deep profiles permit accurate quantification of inherited 10Be concentrations in fan sediment, but restrict the fan age and erosion rate only to a family of permissible age/erosion rate pairs. We use U-series minimum ages and soil observations to constrain the acceptable model solutions, as suggested by Balco et al. (Goldschmidt Conf., 2009). The Imperial Valley fan yields acceptable model fits to 10Be data with predicted erosion of 50 – 140 cm, more than inferred from geomorphic indicators. At the Anza Borrego fan, excess scatter of the 10Be data suggests failure of one or more model assumptions, most likely due to vertical mixing, non-uniform inheritance, or non-representative sampling. Allowing for modest “geologic scatter” (i.e., ~2X analytical errors) and limiting erosion to <5 cm (based on the relict A-layer), we obtain acceptable model fits resulting in ages 0-5 ka older than the U-series age. We conclude that: (1) without additional constraints, a 2 m 10Be depth profile is not sufficient to simultaneously determine both fan age and surface erosion rate, (2) combining 10Be depth profile models with U-series pedogenic carbonate ages and soils observations may increase precision and accuracy in ages for alluvium, and (3) in one case, our approach predicts more erosion than expected from geomorphology.