CO2transport uncertainties from the uncertainties in meteorological fields

摘要: [1] Inference of surface CO2 fluxes from atmospheric observations requires information about large-scale transport and turbulent mixing in the atmosphere, so errors statistics have significant impact on flux estimation. In this paper, we assimilate raw meteorological every 6 hours into a general circulation model with prognostic carbon cycle (CAM3.5) using Local Ensemble Transform Kalman Filter (LETKF) to produce an ensemble analyses that represent best approximation its uncertainty. We quantify uncertainties resulting fields by running forecasts within LETKF-CAM3.5 system forced prescribed fluxes. show are largest over tropical land areas large fossil fuel emissions, between 1.2 3.5 ppm at 0.8 1.8 column-integrated (with OCO-2-like averaging kernel) these regions. further current practice single field has weaker vertical stronger gradient when compared mean initialized fields, especially areas. The magnitude difference can be up 1.5 ppm.