Skip to Content

Quantifying the impact of model bias in convective transport on inferred CO source estimates using multi-spectral CO retrievals from MOPITT

Inverse modeling of atmospheric CO to quantify emissions of CO is sensitive to systematic model errors. Errors in moist convective transport, in particular, are a significant source of model bias. We examine here the impact of such biases on inferred CO source estimates using the GEOS-Chem global chemical transport model and observations of CO from the MOPITT instrument.  We take advantage of the newly available multi-spectral retrievals of CO from MOPITT, which uses information in the near infrared and thermal infrared to provide sensitivity to atmospheric CO throughout the free troposphere and the boundary layer. Using the 4-dimensional variational data assimilation system in GEOS-Chem, we compare the CO source estimates obtained using the profile retrievals, the boundary layer retrievals, and the retrieved CO column abundances from MOPITT. We show that in regions of strong convection, the different vertical information from the boundary layer and middle and upper tropospheric retrievals provides a useful means for quantifying the impact of systematic model errors in convective transport on the inferred CO source estimates.