Sulfur hexafluoride emissions – critical remarks on inversion techniques and inverse model results for the past two decades

Inverse modeling provides a powerful tool to verify national greenhouse gas (GHG) emission inventories by using atmospheric observations. Often, inversions are based on Lagrangian Particle Dispersion Model simulations, where virtual particles are released from observation sites and traced backwards in time to establish a relationship between atmospheric concentrations and emission sources within the simulation period. The fact, that this simulation period is limited due to computational costs, raises two essential questions: (i) How to best define a baseline, that accounts for all emissions that occur prior to the simulation period? (ii) Which period length should be chosen for the backward-simulation?

We show that often used statistical baseline methods have large problems and present a superior global-distribution-based (GDB) approach, that is consistent with the backward-simulation period, accounts for meteorological variability, and leads to inversion results that agree well with known global emission estimates. Our results further show, that longer backward-simulation periods beyond the often used 5 to 10 days increase the correlation between modeled and observed concentrations, and lead to more robust inversion results. Furthermore, they can help to better constrain emissions in regions poorly covered by the observation network.

Based on these methodological results, we perform inversions for sulfur hexafluoride (SF₆) - the most potent GHG regulated under the Kyoto Protocol with an estimated atmospheric lifetime of 3200 years. The inversions are based on 50-days backward-simulations, in-situ and flask measurements from various observation networks, and the GDB baseline method, to achieve globally and regionally consistent SF₆ emissions from 2005 to present.