Growth in NO_x emissions from power plants in China: bottom-up estimates and satellite observations

Abstract. Using OMI (Ozone Monitoring Instrument) tropospheric NO₂ columns and a nested-grid 3-D global chemical transport model (GEOS-Chem), we investigated the growth in NO_x emissions from coal-fired power plants and their contributions to the growth in NO₂ columns in 2005–2007 in China. We first developed a unit-based power plant NO_x emission inventory for 2005–2007 to support this investigation. The total capacities of coal-fired power generation have increased by 48.8% in 2005–2007, with 92.2% of the total capacity additions coming from generator units with size ≥300 MW. The annual NO_x emissions from coal-fired power plants were estimated to be 8.11 Tg NO₂ for 2005 and 9.58 Tg NO₂ for 2007, respectively. The modeled summer average tropospheric NO₂ columns were highly correlated (R² = 0.79–0.82) with OMI measurements over grids dominated by power plant emissions, with only 7–14% low bias, lending support to the high accuracy of the unit-based power plant NO_x emission inventory. The ratios of OMI-derived annual and summer average tropospheric NO₂ columns between 2007 and 2005 indicated that most of the grids with significant NO₂ increases were related to power plant construction activities. OMI had the capability to trace the changes of NO_x emissions from individual large power plants in cases where there is less interference from other NO_x sources. Scenario runs from GEOS-Chem model suggested that the new power plants contributed 18.5% and 10% to the annual average NO₂ columns in 2007 in Inner Mongolia and North China, respectively. The massive new power plant NO_x emissions significantly changed the local NO₂ profiles, especially in less polluted areas. A sensitivity study found that changes of NO₂ shape factors due to including new power plant emissions increased the summer average OMI tropospheric NO₂ columns by 3.8–17.2% for six selected locations, indicating that the updated emission information could help to improve the satellite retrievals.