Articles | Volume 9, issue 10
Research article
24 Oct 2016
Research article |  | 24 Oct 2016

Inverse modelling of NOx emissions over eastern China: uncertainties due to chemical non-linearity

Dasa Gu, Yuhang Wang, Ran Yin, Yuzhong Zhang, and Charles Smeltzer

Abstract. Satellite observations of nitrogen dioxide (NO2) have often been used to derive nitrogen oxides (NOx =  NO + NO2) emissions. A widely used inversion method was developed by Martin et al. (2003). Refinements of this method were subsequently developed. In the context of this inversion method, we show that the local derivative (of a first-order Taylor expansion) is more appropriate than the “bulk ratio” (ratio of emission to column) used in the original formulation for polluted regions. Using the bulk ratio can lead to biases in regions of high NOx emissions such as eastern China due to chemical non-linearity. Inverse modelling using the local derivative method is applied to both GOME-2 and OMI satellite measurements to estimate anthropogenic NOx emissions over eastern China. Compared with the traditional method using bulk ratio, the local derivative method produces more consistent NOx emission estimates between the inversion results using GOME-2 and OMI measurements. The results also show significant changes in the spatial distribution of NOx emissions, especially over high emission regions of eastern China. We further discuss a potential pitfall of using the difference of two satellite measurements to derive NOx emissions. Our analysis suggests that chemical non-linearity needs to be accounted for and that a careful bias analysis is required in order to use the satellite differential method in inverse modelling of NOx emissions.

Short summary
We show in this study that the non-linearity of NOx chemistry implies that the local derivative (of a first-order Taylor expansion) is better suited in the inverse modelling of NOx emissions over eastern China. We used single-grid-cell-based perturbation sensitivity calculation. The method leads to smaller NOx emission over highly polluted regions and higher NOx emission over rural regions. There are also more consistent results from using two different satellite instruments.