Evaluation of pollutant outflow and CO sources during TRACE-P using model-calculated, aircraft-based, and MOPITT-derived CO concentrations

Outflow of CO from Asia during March 2001 was evaluated using data from the TRACE-P mission and the MOPITT instrument in conjunction with model-calculated CO from the University of Maryland CTM. Comparison of model-calculated CO with aircraft measurements was encouraging. Temporal and spatial variations in CO were well captured (mean correlation coefficient of 0.78); however, model-calculated mixing ratios were lower than observed especially below 850 hPa where negative biases of ~60 ppbv were seen. "Non-standard" aerosol effects explain a portion of the negative biases. Below 300 hPa, biases between model-calculated and observed CO distributions were reduced by 6-9 ppbv when OH concentrations were adjusted to take into the account the effect of aerosols on photolysis- and reaction-rates. Inverse modeling of CO was used to estimate the corrections to Asian CO emissions suggested by the combination of model simulations and aircraft measurements in the TRACE-P region. Resulting Asian correction factors were 1.70 +/- 0.36 for fossil fuel/biofuel (ff/bf) emissions and 1.00 +/- 0.30 for bb emissions. Resulting ff/bf emissions were 29.6 +/- 6.2 Tg for March 2001 (323 +/- 67 Tg for an entire year). Resulting bb emissions for March 2001 were 18.1 +/- 5.4 Tg. Comparison of model-calculated CO with MOPITT measurements supported the results from our inverse modeling study. Without exception, mean March 2001 model-calculated CO profiles in the TRACE-P region from a simulation with adjusted CO sources were within a standard deviation of mean March 2001 MOPITT-sampled profiles. Finally, sampling biases need to be considered when interpreting low-latitude MOPITT profiles