Adjoint inverse modeling of NOx emissions over eastern China using satellite observations of NO2 vertical column densities

A four-dimensional variational data assimilation system for optimization of NOx emissions (RC4-NOx) was developed. A parameterized NOx chemistry scheme was introduced into the RC4-NOx system, and key parameters such as chemical production and loss terms of NOx were calculated in advance using the Community Multiscale Air Quality (CMAQ) modeling system. RC4-NOx was applied to optimize NOx emissions over eastern China (EC) in July 1996, 1999, and 2002 using Global Ozone Monitoring Experiment (GOME) satellite observations of NO2 vertical column densities (VCDs) and a priori emissions from the Regional Emission Inventory in Asia (REAS). After assimilation, RC4-NOx generally reproduced the spatial distribution, regional averaged values, and time evolution of GOME NO2 VCDs. Over EC, a priori emissions were reduced by 20% in 1996 and by 8% in 1999, whereas a posteriori emissions were almost the same as a priori emissions in 2002. A priori emissions in the Beijing region were reduced by optimization over the whole simulation period. A posteriori emissions over the Yangtze Delta were larger than a priori emissions in 2002, although they were smaller in both 1996 and 1999. As in other areas, a priori emissions over the North China Plain were reduced in 1996; but those over the eastern part of the plain were increased in 1999, and the area of increased emissions moved slightly westward in 2002. In each region, the growth rates of a posteriori emissions during both 1996–1999 and 1999–2002 became generally larger than those of a priori emissions, and the trends of a posteriori emissions became similar to those of GOME NO2 VCDs. Our inverse modeling analysis indicates that the rate of increase of NOx emissions over EC from 1996 to 2002 was much larger for a posteriori emissions (49%) than for a priori emissions (19%).