A regional data assimilation system for estimating CO surface flux from atmospheric mixing ratio observations-a case study of Xuzhou, China.

Abstract

Carbon monoxide (CO) emission inventory data are crucial for air quality control. However, the emission inventories are labor-intensive and time-consuming and generally have large uncertainties. In this study, we developed a new regional data assimilation system (TracersTracker) for estimating the surface CO emission flux from continuous mixing ratio observations using the proper orthogonal decomposition (POD)-based four-dimensional variational (4D-VAR) data assimilation method (POD-4DVar) and a coupled regional model (Weather Research and Forecasting model (WRF) with the Models-3 Community Multi-scale Air Quality (CMAQ) model). This system was applied to estimate CO emissions in Xuzhou city, China. An experiment was conducted with the continuous hourly surface CO mixing ratio observations from 21 monitoring towers in January and July of 2016. The experimental results of the system were examined and compared with the continuous surface CO observations (a priori emission). We found that the retrieved CO emission fluxes were higher than the a priori emission and were mainly distributed in urban and industrial areas, which were 104% higher in January (winter) and 44% higher in July (summer).