Soil CO2 efflux simulations using Monte Carlo method and implications for recording paleo-atmospheric PCO2 in pedogenic gibbsite

Monte Carlo simulation testing of the sensitivity of the Fickian diffusion model used for paleo-PCO2 proxies indicates that assuming annual average soil respiration rate results in over-estimation of PCO2. In all cases, the natural variability of the factors of the model, including soil porosity and inverse tortuosity and the delta C-13 value of the soil produced CO2 were not able to account for discrepancies between measured delta C-13 values from a modern soil and modeled results using established assumptions. Assuming that there is negligible carbon isotope fractionation during the occlusion of CO2 in pedogenic gibbsite, the flaw in the model appears to be the assumed value of soil respiration rate. If correct, the implication is that previous paleo-PCO2 estimates from soil hydroxides may be too high by as much as a factor of ten. (C) 2011 Elsevier B.V. All rights reserved.