Articles online

Ozone in China: Spatial Distribution and Leading Meteorological Factors Controlling O3 in 16 Chinese Cities

Category: MAPS: Spatial Variation - China

Volume: 18 | Issue: 9 | Pages: 2287-2300
DOI: 10.4209/aaqr.2017.10.0368
PDF | Supplemental material

Export Citation:  RIS | BibTeX

Xi Gong1,2, Song Hong 1,3, Daniel A. Jaffe 2,4

  • 1 School of Resource and Environmental Sciences, Wuhan University, Wuhan 430079, China
  • 2 School of Science, Technology, Engineering and Mathematics, University of Washington Bothell, Bothell, WA 98011, USA
  • 3 Shenzhen Research Institute, Wuhan University, Shenzhen 518000, China
  • 4 Department of Atmospheric Sciences, University of Washington Seattle, Seattle, WA 98195, USA


We summarized 3 seasonal patterns of O3 for 16 cities including 7 megacities.
GAM is a useful tool to examine daily O3 with meteorological factors (R2 = 0.43–0.9).
Top 3 leading factors were identified in 16 cities using F statistic value of GAM.
Wind directions affect O3 for 3 coastal cities and Beijing categorized by HYSPLIT.
O3 in Beijing “Parade Blue” period is mainly controlled by meteorological factors.


Tropospheric ozone (O3) is one of the major air pollutants in China. This paper examined the O3 concentration in 16 important Chinese cities including 7 megacities and developed a statistical model named Generalized Additive Model (GAM) as a function of different factors to estimate the maximum daily 8 h (MDA8) O3 during 2014–2016 and how the leading factors impacts O3. We found that: (1) Three seasonal patterns of O3 have been summarized in the spatial-temporal analysis and summer is the highest season in most of the cities. (2) GAM performs very well that it can capture 43–90% of daily O3 variations. (3) DOY (day of year) and 6 meteorological factors of daily average relative humidity at 1000 mb, daily maximum temperature at 2 m, daily average zonal wind speed at 700 mb, distance of trajectory back 12-hour, surface pressure and geopotential height at 500 mb are sensitive for all 16 cities. The sequence of the leading factors is the same in each group respectively (3 group categories: Beijing, Shijiazhuang and Kunming; Harbin, Hohhot and Dalian; Chengdu and Wuhan). The other 8 cities have different leading factor combination. (4) HYSPLIT back trajectory data can help us to know the importance of transport direction for O3 concentration in Beijing and other three coastal cities Dalian, Shanghai and Guangzhou. (5) During the Beijing “Parade Blue” period in the summer of 2015, NO2 was reduced by 44.6% but O3 was only reduced by 15.7%. Most of these O3 changes can be explained by meteorological variations such as wind direction and air temperature.


Ozone Spatial-temporal distribution Meteorology Generalized Additive Model China

Related Article

The Controlling Factors of Photochemical Ozone Production in Seoul, South Korea

Saewung Kim , Daun Jeong, Dianne Sanchez, Mark Wang, Roger Seco, Donald Blake, Simone Meinardi, Barbara Barletta, Stacey Hughes, Jinsang Jung, Deugsoo Kim, Gangwoong Lee, Meehye Lee, Joonyoung Ahn, Sang-Deok Lee, Gangnam Cho, Min-Young Sung, Yong-Hwan Lee, Rokjin Park