2013–2019 increases of surface ozone pollution in China: anthropogenic and meteorological influences

Surface ozone data from the Chinese Ministry of Ecology and Environment (MEE) network show sustained increases across the country over the 2013–2019 period. Despite Phase 2 of Clean Air Action targeting ozone pollution, ozone was higher in 2018–2019 than in previous years. The mean summer 2013–2019 trend of maximum 8-h average 15 (MDA8) ozone was 1.9 ppb a across China and 3.3 ppb a in the North China Plain (NCP). Fitting ozone to meteorological variables with a multiple linear regression model shows that meteorology played a significant but not dominant role in the 2013–2019 ozone trend, contributing 0.70 ppb a across China and 1.4 ppb a in the NCP. Higher June-July temperatures over the NCP were the main meteorological driver, particularly in recent years (2017–2019), and were associated with increased foehn winds. NCP data for 2017–2019 show a 15% continuing decrease in fine particulate 20 matter (PM2.5) and flat emissions of volatile organic compounds (VOCs), which would explain the continued anthropogenic increase in ozone. VOC emission controls, as targeted by Phase 2 of the Chinese Clean Air Action, are needed to reverse the increase of ozone. 25 https://doi.org/10.5194/acp-2020-298 Preprint. Discussion started: 16 April 2020 c © Author(s) 2020. CC BY 4.0 License.


Introduction
Surface ozone is a serious air pollution issue over much of eastern China (Ma et al., 2012;. Measurements from the Chinese Ministry of Environment and Ecology (MEE) network of sites frequently exceed the national air quality standard of 160 µg m -3 , corresponding to 82 ppb at 298 K and 1013 hPa (Li et al., 2017;Shen et al., 2019a;Fan et al., 2020). The Clean Air Action initiated in 2013 imposed rapid decreases in pollutant emissions (Chinese State Council, 5 2013) and resulted in large decreases in fine particulate matter (PM2.5) concentrations (Zhai et al., 2019;. However, ozone increased by 1-3 ppb a -1 over the 2013-2017 period in megacity clusters of eastern China (Lu et al., 2018;Li et al., 2019a;Lu et al. 2020), partly offsetting the health benefits from improved PM2. 5 (Dang and Liao, 2019;. Phase 2 of Clean Air Action starting in 2018 (Chinese State Council, 2018) imposed new emission controls targeted at ozone. Here we show that the increasing ozone trend in eastern China has continued 10 through 2019, driven by both anthropogenic emission and meteorological trends, and stressing the urgent need for more vigorous emission controls.
Ozone in polluted regions is produced by photochemical reactions of volatile organic compounds (VOCs) and nitrogen oxides (NOx ≡ NO + NO2), enabled by hydrogen oxide radicals (HOx ≡OH + peroxy radicals) as oxidants. VOCs and NOx are emitted by fuel combustion, and VOCs have additional industrial sources (Zheng et al., 2018) and biogenic 15 sources (Guenther et al., 2012). HOx is produced photochemically from ozone and water, formaldehyde (HCHO), nitrous acid, and other precursors (Tan et al., 2019). Ozone is highest in summer when photochemistry is most active . Meteorological conditions play an important role in modulating ozone concentrations, not only through transport but also by affecting natural emissions and chemical rates (Jacob and Winner, 2009;Shen et al., 2016;Lu et al., 2019). 20 A number of studies have investigated the roles of anthropogenic and meteorological factors in driving the 2013-2017 ozone trend, and concluded that meteorological factors were not negligible but anthropogenic factors were dominant Li et al., 2019a;Yu et al., 2019;Liu et al., 2020). Our previous work (Li et al., 2019a(Li et al., , 2019b found that the decrease of PM2.5 was a major factor driving the increase of ozone due to the role of PM2.5 as scavenger of hydroperoxy (HO2) radicals and NOx that would otherwise produce ozone. Here we extend the analysis of 25 ozone trends to 2019, into the implementation of Clean Air Action Phase 2, and bring in satellite and ground-based observations to relate the most recent ozone trends to those of VOC (Shen et al., 2019b) and NOx emissions (Zheng et al., 2018;Shah et al., 2020).

Surface measurements
Hourly concentrations of ozone, PM2.5, and NO2 are taken from the MEE website (http: //106.37.208.233:20035)  access: 20 July 2020). The TROPOMI HCHO data are freely accessed from https://s5phub.copernicus.eu/dhus/ (last access: 28 February 2020) and we only use observations with quality assurance value larger than 0.5. This filter effectively removes data with cloud fraction larger than 0.5. Interannual trends in HCHO columns could be affected by temperature-dependent emissions of biogenic VOCs (Palmer et al., 2006). Following Zhu et al. (2017, we remove this contribution by regressing JJA monthly mean HCHO columns onto noon (13:00 local time) surface air temperatures, and 25 then subtracting this fitted temperature dependency.

Stepwise multiple linear regression (MLR) model
To quantify the role of meteorology in driving 2013-2019 ozone trends, we use the same stepwise multiple linear regression (MLR) modeling approach as Li et al. (2019a). This modeling approach relates the month-to-month variability of MDA8 ozone to that of meteorological variables. Consistent meteorological fields for 2013-2019 were obtained from the NASA Modern-Era Retrospective Analysis for Research and Applications, Version 2 (MERRA-2) product (https://gmao.gsfc.nasa.gov/reanalysis/MERRA-2, last access: 28 February 2020) (Gelaro et al., 2017). The MERRA-2 data have a spatial resolution of 0.5° latitude × 0.625° longitude. We average the daily MDA8 ozone from the MEE network onto the MERRA-2 grid. Firstly, the regression model is applied to select the key meteorological parameters 5 driving the day-to-day variability of ozone for each grid cell. There are nine MERRA-2 meteorological variables considered as ozone covariates, including daily maximum 2-m air temperature (Tmax), 10-m zonal wind (U10) and meridional wind (V10), boundary layer height (PBLH), total cloud area fraction (TCC), rainfall (Rain), sea level pressure (SLP), relative humidity (RH), and 850-hPa meridional wind (V850), following (Li et al., 2019a). The meteorology fields are averaged over 24 h for use in the MLR model except for PBLH and TCC, which are averaged over daytime hours (8-10 20 local time), and for Tmax (daily maximum).
Secondly, to avoid overfitting, only the three locally dominant meteorological parameters are regressed onto the deseasonalized monthly MDA8 ozone to fit the role of 2013-2019 meteorological variability. The top three variables are selected based on their individual contribution to the regressed ozone, along with the requirement that they are statistically significant above the 95% confidence level in the MLR model. They will differ for each 0.5° × 0.625° grid 15 cell. We show these top three meteorological drivers for ozone variability in Figure S1-S3 for different locations in China.
Thirdly, we fit the observed monthly ozone anomalies by applying these dominant meteorological drivers in the MLR model. The coefficients of determination (R 2 ) for the MLR model are generally above 0.4-0.5 for polluted regions of China which are of most interest to us ( Figure S4). Remote locations with background ozone levels have less ozone 20 variability and are thus harder to fit. Similar MLR models have been extensively employed to quantify the effect of meteorological variability on air pollutants (e.g., Tai et al., 2010;Otero et al., 2018;Zhai et al., 2019;Han et al., 2020).
Finally, the trend in regressed ozone is taken to reflect the meteorological contribution, and the residual is then taken to reflect the presumed anthropogenic contribution, with the statistical significance of the anthropogenic trend determined by Student's t-test. We have followed this approach before to isolate the anthropogenic trends of ozone and PM2.5 (Li et 25 al., 2019a;Zhai et al., 2019). Similar statistical decomposition of anthropogenic and meteorological contributions to air pollutant trends has been also employed by previous studies (e.g., Chen et al. 2019;Yu et al., 2019;. The effect of biogenic VOCs on ozone trends depends on meteorological and land cover drivers. Meteorological drivers, in particular temperature, would be accounted for in the MLR model. The effect of land cover changes is expected to be small over the 7-year time horizon of our analysis (Fu and Tai, 2014) We first present the general 2013-2019 summer ozone trends in China and their statistically decomposed meteorological and anthropogenic contributions. Ozone trends over the major megacity clusters in China are highlighted. We go on to more specifically attribute the meteorological and anthropogenic drivers of recent ozone trends over the North China Plain, where the ozone increase is the highest.  restricted to the Shandong Peninsula and northeastern China (including Heilongjiang, Jilin, and Liaoning provinces). The mean trend for China is 1.9 ppb a -1 (p<0.01). Trends in the four megacity clusters are 3.3 ppb a -1 (p<0.01) for NCP, 1.6 ppb a -1 (p<0.01) for YRD, 1.1 ppb a -1 (p=0.03) for PRD, and 0.7 ppb a -1 (p=0.23) for SCB (Table 1). The increases are largest in the NCP, which could be explained by greater influence of radical scavenging by PM2.5 (Li et al., 2019a, 25 2019b).

Figure 2b
shows the meteorologically driven ozone trends, as determined by fitting ozone to meteorological variables with the MLR model. We find an average meteorologically driven trend of 0.7 ppb a -1 (p<0.01) for China. Ozone trends over 2013-2019 in the NCP and PRD are significantly contributed by meteorology, and this is particularly driven by 2018-2019 (Table 1). Similar to our previous study for 2013-2017 (Li et al., 2019a), the most important meteorological 6 predictor variables in the MLR model are daily maximum temperature for the NCP and meridional wind at 850 hPa for the PRD ( Figure S1). These dominant meteorological parameters are also consistent with the findings from other studies (Gong and Liao, 2019;Han et al., 2020). Hot weather is the main meteorological driver for high ozone in the NCP, and we will elaborate on this in the next section. The main meteorological driver for the ozone increase in the PRD is the weakening of the summer monsoonal flow (Figure 3) that ventilates the PRD with marine air. 5 On the other hand, we find that meteorology mitigated ozone pollution increases over northeastern China and the Shandong Peninsula. As shown in Figure S2