Articles | Volume 22, issue 21
Research article
03 Nov 2022
Research article |  | 03 Nov 2022

Inverse modelling of Chinese NOx emissions using deep learning: integrating in situ observations with a satellite-based chemical reanalysis

Tai-Long He, Dylan B. A. Jones, Kazuyuki Miyazaki, Kevin W. Bowman, Zhe Jiang, Xiaokang Chen, Rui Li, Yuxiang Zhang, and Kunna Li

Data sets

Chemical Reanalysis Products K. Miyazaki, K. Bowman, T. Sekiya, H. Eskes, F. Boersma, H. Worden, N. Livesey, V. H. Payne, K. Sudo, Y. Kanaya, M. Takigawa, and K. Ogochi

ERA5 hourly data on single levels from 1959 to present H. Hersbach, B. Bell, P. Berrisford, G. Biavati, A. Horányi, J. Muñoz Sabater, J. Nicolas, C. Peubey, R. Radu, I. Rozum, D. Schepers, A. Simmons, C. Soci, D. Dee, and J.-N. Thépaut

Model settings and surface measurements for air quality study Zhe Jiang

Model code and software

Unet_Chinese_NOx Tai-Long He

Short summary
We use a deep-learning (DL) model to estimate Chinese NOx emissions by combining satellite analysis and in situ measurements. Our results are consistent with conventional analyses of Chinese NOx emissions. Comparison with mobility data shows that the DL model has a better capability to capture changes in NOx. We analyse Chinese NOx emissions during the COVID-19 pandemic lockdown period. Our results illustrate the potential use of DL as a complementary tool for conventional air quality studies.
Final-revised paper