Coal-derived rates of atmospheric dust deposition during the Permian

Despite widespread evidence for atmospheric dust deposition prior to the Quaternary, quantitative rate data remains sparse. As dust influences both climate and biological productivity, the absence of quantitative dust data limits the comprehensiveness of models of pre-Quaternary climate and biogeochemical cycles. Here, we propose that inorganic matter contained in coal primarily records atmospheric dust deposition. To test this, we use the average concentration of inorganic matter in Permian coal to map global patterns and deposition rates of atmospheric dust over Pangea. The dust accumulation rate is calculated assuming Permian peat carbon accumulation rates in temperate climates were similar to Holocene rates and accounting for the loss of carbon during coalification. Coal-derived rates vary from 0.02 to 25 g m(-2) year(-1), values that fall within the present-day global range. A well-constrained East-West pattern of dust deposition corresponding to expected palaeoclimate gradients extends across Gondwana with maximum dust deposition rates occurring close to arid regions. A similar pattern is partially defined over the northern hemisphere. Patterns are consistent with the presence of two large global dust plumes centred on the tropics. The spatial patterns of dust deposition were also compared to dust cycle simulations for the Permian made with the Community Climate System Model version 3 (CCSM3). Key differences between the simulations and the coal data are the lack of evidence for an Antarctic dust source, higher than expected dust deposition over N and S China and greater dust deposition rates over Western Gondwana. This new coal-based dust accumulation rate data expands the pre-Neogene quantitative record of atmospheric dust and can help to inform and validate models of global circulation and biogeochemical cycles over the past 350 Myr. (C) 2015 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved.