Mid-summer aircraft observations suggest a net biological source of CO2 to the atmosphere in the Gulf coast states. These observations are inconsistent with terrestrial biosphere models that show strong net uptake of CO2 in this region in summer. Methane observations downwind of major sources in the MidAtlantic region suggest that these sources are represented fairly well by existing emissions inventories. Flux estimation in other regions and seasons is underway.
Spatially-coherent differences in GHGs that extend throughout the depth of the troposphere are observed at frontal boundaries in summer and winter. These spatial structures are captured in global and mesoscale model simulations, though the simulated GHG mole fractions are sometimes biased with respect to observations, suggesting potential biases in synoptic transport. Mesoscale simulations of atmospheric CO2 overestimate spatial differences in ABL CO2 mole fractions in fair weather conditions as compared to airborne observations and the CarbonTracker global inverse modeling system. ABL depths appear to be simulated fairly well by both mesoscale and global modeling systems, suggesting that either weather-scale flux amplitudes are overestimated by CarbonTracker, or the mesoscale model is lacking parameterized transport above the ABL.
Evaluation of the spatial variability in both OCO-2 and airborne lidar XCO2 observations is defining the regional precision and accuracy of these observational methods.
These findings are moving us toward improved regional GHG inverse flux estimates via a synthesis of better understanding of prior fluxes, atmospheric transport, and satellite CO2 observations.