Abstract

Modelled concentrations of gases and aerosols are the result of a myriad of processes, such as emissions and deposition at the surface, transport and coupled nonlinear chemistry. Detecting a change and quantifying emissions from observed concentrations can be challenging because of biases in the ... Show moreModelled concentrations of gases and aerosols are the result of a myriad of processes, such as emissions and deposition at the surface, transport and coupled nonlinear chemistry. Detecting a change and quantifying emissions from observed concentrations can be challenging because of biases in the prior set of model inputs and parameters. In addition to the variety of sources, anthropogenic bottom-up emissions are defined for different sectors (industry, transportation, energy, residential) that are difficult to constrain in the inverse-modeling process because of the underdetermined nature of the problem. From this point of view, the COVID-19 lockdown containment provides a real-world experiment where some specific sectors, such as transportation, were almost completely shut down. We use daily sectoral bottom-up emissions from the Copernicus Atmosphere Monitoring Service (CAMS) as a baseline to compare with a bottom-up scenario with reduction factors applied to different sectors. We compare the results with top-down estimates of daily reduction factors estimated from the assimilation of TROPOMI (TROPOspheric Monitoring Instrument) NO2 and SO2 observations. From this comparison, we perturb emissions sectors to define an ensemble of Community Atmosphere Model with chemistry (CAM-chem) simulations. We then investigate the optimization of sectoral and daily emissions, and the concentration of carbon monoxide (CO) from the assimilation of satellite retrievals. Observations that are assimilated include the Measurements of the Pollution In The Troposphere (MOPITT), Cross-track Infrared Sounder (CrIS) and TROPOMI. Finally, we compare the model results with available in-situ measurements and inverted emissions with other bottom-up estimates. Show less