An attempt at estimating Paris area CO 2 emissions from atmospheric concentration measurements

摘要: Abstract. Atmospheric concentration measurements are used to adjust the daily monthly budget of fossil fuel CO2 emissions Paris urban area from prior estimates established by Airparif local air quality agency. Five atmospheric monitoring sites available, including one at top Eiffel Tower. The inversion is based on a Bayesian approach, and relies an transport model with spatial resolution 2 km boundary conditions global coarse grid model. adjusts knowledge about anthropogenic biogenic fluxes inventory ecosystem model, respectively, corrections temporal 6 h, while keeping distribution emission inventory. These assumptions regarding autocorrelation uncertainties within cycle, day day. comparison against simulation driven priori surface shows significant differences upwind area, which suggests large uncertain contribution distant sources sinks variability. This advocates that should aim minimising model–data misfits in upwind–downwind gradients rather than mole fractions individual sites. Another conclusion direct model–measurement variability Tower poorly represented for most wind speeds directions. model's inability reproduce heart city makes such ill-suited inversion. need constrain budgets whole assimilation fraction between edge only. significantly improves agreement measured modelled gradients. Realistic retrieved two 30-day periods suggest overestimate AirParif Similar inversions over longer necessary proper evaluation optimised independent data.