Quantification of the Uptake of Anthropogenic Emissions of Atmospheric CO2

We quantify the efficiency of the combined land and ocean carbon sink of anthropogenic CO₂, from 1959 to present, based on analysis of a variety of data records. We assess the temporal trend in the airborne fraction (AF) of CO₂, defined as the annual growth of atmospheric CO₂ (G_ATM) divided by the total annual anthropogenic CO₂ emissions (E_NET). The motivation for this study is the suggestion, in Section 6.4.3.3 of IPCC (2013), that AF has risen in a statistically significant manner since 1959. Our analysis uses a model to quantify the effect on G_ATM from ENSO, volcanic eruptions, and other natural factors. We then quantify the trend in AF based upon two time series of G_ATM: 1) the raw data; 2) a time series of G_ATM adjusted for these natural influences. A Monte-Carlo simulation coupled with an auto-regressive model is used to generate synthetic time series for raw and adjusted G_ATM as well as fossil fuel and land use change components of E_NET. The statistical significance of the change in the efficiency of the combined land and ocean sink is assessed using Gaussian fits to the resulting probability distribution functions of the slopes of time series for both AF_RAW and AF_ADJUSTED. Our preliminary analysis indicates that AF_RAW has risen over time at an amount significant to about the 1σ level and that AF_ADJUSTED has declined slightly, but not in a statistically significant manner. Our study highlights the need to account for natural factors that can strongly influence the atmospheric growth of CO₂ on an annual basis, when assessing long-term change in the global cycle. We will conclude by offering suggestions on how existing data can be used to provide meaningful comparisons to the representation of climate-carbon feedbacks implemented in GCMs that carried out the C4MIP experiment for Phase 6 of the Coupled Model Intercomparison Project Phase 6 (CMIP6) (Eyring et al., GMD, 2016).