Attribution of Air Quality Benefits to Clean Winter Heating Policies in China: Combining Machine Learning with Causal Inference.

Heating is a major source of air pollution. To improve air quality, a range of clean heating policies were implemented in China over the past decade. Here, we evaluated the impacts of winter heating and clean heating policies on air quality in China using a novel, observation-based causal inference approach. During 2015-2021, winter heating causally increased annual PM2.5, daily maximum 8-h average O3, and SO2 by 4.6, 2.5, and 2.3 µg m-3, respectively. From 2015 to 2021, the impacts of winter heating on PM2.5 in Beijing and surrounding cities (i.e., "2 + 26" cities) decreased by 5.9 µg m-3 (41.3%), whereas that in other northern cities only decreased by 1.2 µg m-3 (12.9%). This demonstrates the effectiveness of stricter clean heating policies on PM2.5 in "2 + 26" cities. Overall, clean heating policies caused the annual PM2.5 in mainland China to reduce by 1.9 µg m-3 from 2015 to 2021, potentially avoiding 23,556 premature deaths in 2021.

Impacts of meteorology and precursor emission change on O3 variation in Tianjin, China from 2015 to 2021.

Deterioration of surface ozone (O3) pollution in Northern China over the past few years received much attention. For many cities, it is still under debate whether the trend of surface O3 variation is driven by meteorology or the change in precursors emissions. In this work, a time series decomposition method (Seasonal-Trend decomposition procedure based on Loess (STL)) and random forest (RF) algorithm were utilized to quantify the meteorological impacts on the recorded O3 trend and identify the key meteorological factors affecting O3 pollution in Tianjin, the biggest coastal port city in Northern China. After "removing" the meteorological fluctuations from the observed O3 time series, we found that variation of O3 in Tianjin was largely driven by the changes in precursors emissions. The meteorology was unfavorable for O3 pollution in period of 2015-2016, and turned out to be favorable during 2017-2021. Specifically, meteorology contributed 9.3 µg/m3 O3 (13%) in 2019, together with the increase in precursors emissions, making 2019 to be the worst year of O3 pollution since 2015. Since then, the favorable effects of meteorology on O3 pollution tended to be weaker. Temperature was the most important factor affecting O3 level, followed by air humidity in O3 pollution season. In the midday of summer days, O3 pollution frequently exceeded the standard level (>160 µg/m3) at a combined condition with relative humidity in 40%-50% and temperature > 31°C. Both the temperature and the dryness of the atmosphere need to be subtly considered for summer O3 forecasting.

Understanding Sources and Drivers of Size-Resolved Aerosol in the High Arctic Islands of Svalbard Using a Receptor Model Coupled with Machine Learning.

Atmospheric aerosols are important drivers of Arctic climate change through aerosol-cloud-climate interactions. However, large uncertainties remain on the sources and processes controlling particle numbers in both fine and coarse modes. Here, we applied a receptor model and an explainable machine learning technique to understand the sources and drivers of particle numbers from 10 nm to 20 µm in Svalbard. Nucleation, biogenic, secondary, anthropogenic, mineral dust, sea salt and blowing snow aerosols and their major environmental drivers were identified. Our results show that the monthly variations in particles are highly size/source dependent and regulated by meteorology. Secondary and nucleation aerosols are the largest contributors to potential cloud condensation nuclei (CCN, particle number with a diameter larger than 40 nm as a proxy) in the Arctic. Nonlinear responses to temperature were found for biogenic, local dust particles and potential CCN, highlighting the importance of melting sea ice and snow. These results indicate that the aerosol factors will respond to rapid Arctic warming differently and in a nonlinear fashion.

Marginal reduction in surface NO2 attributable to airport shutdown: A machine learning regression-based approach.

Emissions from aviation and airport-related activities degrade surface air quality but received limited attention relative to regular transportation sectors like road traffic and waterborne vessels. Statistically, assessing the impact of airport-related emissions remains a challenge due to the fact that its signal in the air quality time series data is largely dwarfed by meteorology and other emissions. Flight-ban policy has been implemented in a number of cities in response to the COVID-19 spread since early 2020, which provides an unprecedented opportunity to examine the changes in air quality attributable to airport closure. It would also be interesting to know whether such an intervention produces extra marginal air quality benefits, in addition to road traffic. Here we investigated the impact of airport-related emissions from a civil airport on nearby NO2 air quality by applying machine learning predictive model to observational data collected from this unique quasi-natural experiment. The whole lockdown-attributable change in NO2 was 16.7 µg/m3, equals to a drop of 73% in NO2 with respect to the business-as-usual level. Meanwhile, the airport flight-ban aviation-attributable NO2 was 3.1 µg/m3, accounting for a marginal reduction of 18.6% of the overall NO2 change that driven by the whole lockdown effect. The airport-related emissions contributed up to 24% of the local ambient NO2 under normal conditions. Additionally, the average impact of airport-related emissions on the nearby air quality was ∼0.01 ± 0.001 µg/m3 NO2 per air-flight. Our results highlight that attention needs to be paid to such a considerable emission source in many places where regular air quality regulatory measures were insufficient to bring NO2 concentration into compliance with the health-based limit.

Spring Festival and COVID-19 Lockdown: Disentangling PM Sources in Major Chinese Cities.

Responding to the 2020 COVID-19 outbreak, China imposed an unprecedented lockdown producing reductions in air pollutant emissions. However, the lockdown driven air pollution changes have not been fully quantified. We applied machine learning to quantify the effects of meteorology on surface air quality data in 31 major Chinese cities. The meteorologically normalized NO2, O3, and PM2.5 concentrations changed by -29.5%, +31.2%, and -7.0%, respectively, after the lockdown began. However, part of this effect was also associated with emission changes due to the Chinese Spring Festival, which led to ∼14.1% decrease in NO2, ∼6.6% increase in O3 and a mixed effect on PM2.5 in the studied cities that largely resulted from festival associated fireworks. After decoupling the weather and Spring Festival effects, changes in air quality attributable to the lockdown were much smaller: -15.4%, +24.6%, and -9.7% for NO2, O3, and PM2.5, respectively.

Impact of meteorological condition changes on air quality and particulate chemical composition during the COVID-19 lockdown.

Stringent quarantine measures during the Coronavirus Disease 2019 (COVID-19) lockdown period (January 23, 2020 to March 15, 2020) have resulted in a distinct decrease in anthropogenic source emissions in North China Plain compared to the paralleled period of 2019. Particularly, 22.7% decrease in NO2 and 3.0% increase of O3 was observed in Tianjin, nonlinear relationship between O3 generation and NO2 implied that synergetic control of NOx and VOCs is needed. Deteriorating meteorological condition during the COVID-19 lockdown obscured the actual PM2.5 reduction. Fireworks transport in 2020 Spring Festival (SF) triggered regional haze pollution. PM2.5 during the COVID-19 lockdown only reduced by 5.6% in Tianjin. Here we used the dispersion coefficient to normalize the measured PM2.5 (DN-PM2.5), aiming to eliminate the adverse meteorological impact and roughly estimate the actual PM2.5 reduction, which reduced by 17.7% during the COVID-19 lockdown. In terms of PM2.5 chemical composition, significant NO3- increase was observed during the COVID-19 lockdown. However, as a tracer of atmospheric oxidation capacity, odd oxygen (Ox = NO2 + O3) was observed to reduce during the COVID-19 lockdown, whereas relative humidity (RH), specific humidity and aerosol liquid water content (ALWC) were observed with noticeable enhancement. Nitrogen oxidation rate (NOR) was observed to increase at higher specific humidity and ALWC, especially in the haze episode occurred during 2020SF, high air humidity and obvious nitrate generation was observed. Anomalously enhanced air humidity may response for the nitrate increase during the COVID-19 lockdown period.

Size distribution and chemical characteristics of particles from crop residue open burning in North China.

Crop residue open burning is an important emission source of ambient particles in China. This study analyzed the particle emission characteristics of crop residue open burning through combustion experiments with a novel open combustion simulation device using three typical crop straws in north China (corn, wheat, and rice). Particle samples size ranging from 0.006-9.890 µm were collected by an Electrical Low Pressure Impactor plus, a high size-resolution instrument capable of dividing particles into 14 size stages. The size distributions of organic carbon (OC), elemental carbon (EC), water-soluble ions, and elements were analyzed, and source chemical profiles were constructed for PM0.1, PM1, PM2.5, and PM10. The number concentration of particles was concentrated in the Aiken nuclei mode (0.006-0.054 µm), accounting for 75% of the total number, whereas the mass concentration was concentrated in the accumulation mode (0.054-0.949 µm), accounting for 85.43% of the mass loading. OC, EC, Cl-, and K(include total K and water-soluble K) were the major chemical components of the particles, whose mass percentage distributions differed from those of other components. These five main components exhibited a bell-shaped size distribution in the 0.006-9.890 µm range, whereas the other components exhibited a U-shaped distribution. Among the chemical profiles for PM0.1-PM10, OC was the most important component at 10-30%, followed by EC at 2%-8%. The proportions of K+, Cl-, and K varied substantially in different experimental groups, ranging from 0-15%, and K+ and Cl- were significantly correlated (r = 0.878, α = 0.000).

Dispersion Normalized PMF Provides Insights into the Significant Changes in Source Contributions to PM2.5 after the COVID-19 Outbreak.

Factor analysis utilizes the covariance of compositional variables to separate sources of ambient pollutants like particulate matter (PM). However, meteorology causes concentration variations in addition to emission rate changes. Conventional positive matrix factorization (PMF) loses information from the data because of these dilution variations. By incorporating the ventilation coefficient, dispersion normalized PMF (DN-PMF) reduces the dilution effects. DN-PMF was applied to hourly speciated particulate composition data from a field campaign that included the start of the COVID-19 outbreak. DN-PMF sharpened the morning coal combustion and rush hour traffic peaks and lowered the daytime soil, aged sea salt, and waste incinerator contributions that better reflect the actual emissions. These results identified significant changes in source contributions after the COVID-19 outbreak in China. During this pandemic, secondary inorganic aerosol became the predominant PM2.5 source representing 50.5% of the mean mass. Fireworks and residential burning (32.0%), primary coal combustion emissions (13.3%), primary traffic emissions (2.1%), soil and aged sea salt (1.2%), and incinerator (0.9%) represent the other contributors. Traffic decreased dramatically (70%) compared to other sources. Soil and aged sea salt also decreased by 68%, likely from decreased traffic.

Insight into the critical factors determining the particle number concentrations during summer at a megacity in China.

To identify the critical factors impacting the number concentration of particles with the aerodynamic diameters less than 2.5µm (PNC2.5), the continuous measurement of PNC2.5, chemical components in PM2.5, gaseous pollutants and meteorological conditions were conducted at an urban site in Tianjin in June 2015. Results indicated that the average PNC2.5 was 2839±2430 dN/dlogDp 1/cm3 during the campaign. Compared to other meteorological parameters, the relative humidity (RH) had the strongest relationship with PNC2.5, with a Pearson's correlation coefficient of 0.53, and RH larger than 30% influenced strongly PNC2.5. The important influence of secondary reactions on PNC2.5 was inferred due to higher correlation coefficients between PNC2.5 and SO42-, NO3-, NH4+ (r=0.78-0.89; p<0.01) and between PNC2.5 and ratios that represent the conversion of nitrogen and sulfur oxides to particulate matter (r=0.42-0.49; p<0.01). Under specific RH conditions, there were even stronger correlations between PNC2.5 and NO3-, SO42-, NH4+, while those between PNC2.5 and EC, OC were relatively weak, especially when RH exceeded 50%. Principal component analysis (PCA) and Pearson's correlation analysis indicated that secondary sources, vehicle emission and coal combustion might be major contributors to PNC2.5. Backward trajectory and potential source contribution function (PSCF) analysis suggested that the transport of air masses originated from these regions around Tianjin (Liaoning, Hebei, Shandong and Jiangsu) influenced critically PNC2.5. The north of Jiangsu, the west of Shandong, and the east of Hebei were distinguished as major potential source-areas of PNC2.5 by PSCF model.