This page presents results of inverse modeling of surface sources of carbon monoxide with MOPITT data and global chemistry transport model MOZART 2. This research was sponsored by the MOPITT program and NASA interdisciplinary grant "A study of Tropospheric Budgets of CO and CH4 Using Data Assimilation". A final report to NASA can be found here: EOS IDS Final Report.
The overall technical goal of this study was to apply mathematical techniques of estimation theory to global atmospheric model and observational data to constrain poorly known model parameters (e.g., surface emissions of CO, OH distributions), and to improve our ability to predict the future evolution of the chemical composition of the troposphere. We employ Bayesian inversion approach for deriving surface sources of carbon monoxide from MOPITT data. Inverse modeling of surface emissions is being performed for different seasons with the goal to estimate temporal variability of the surface fluxes.
Check out this article from NASA's Earth observatory to learn more about importance of CO. Click on the left picture below to see an animation of time evotion of CO emitted from different continents. Click on the right picture to see evolution of CO emitted by 50 largest polluters in Coloradon in Dec 2001.
The direct surface emissions of CO were adjusted to minimize the disagreements between the MOPITT data and the distribution of CO computed with the MOZART 2 model. The transport model uses NCEP or ECMWF re-analysis dynamical fields. The chemical production of CO is not optimized. A priori monthly emissions are taken from an early version of EDGAR_v3 and GEIA inventories are used for the inversion.
Five sources of emissions were optimized: biomass burning, anthropogenic, biofuel, biogenic, and oceanic. Emissions for each source were derived for 12 months starting from April 2000 until March 2001.
