Sensitivity of Air Pollution Exposure and Disease Burden to Emission Changes in China Using Machine Learning Emulation. (20th June 2022)

• Record Type: Journal Article
• Authors: Conibear, Luke; Reddington, Carly L.; Silver, Ben J.; Chen, Ying; Knote, Christoph; Arnold, Stephen R.; Spracklen, Dominick V.
• Is Part Of: GeoHealth. Volume 6:Number 6(2022)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-06-20
• Subjects: emulator -- machine learning -- air quality -- health impact assessment -- China -- particulate matter; Environmental health -- Periodicals; Electronic journals; Periodicals; 616.98
• DOI: 10.1029/2021GH000570
• Languages: English
• ISSNs: 2471-1403

Abstract

Abstract: Machine learning models can emulate chemical transport models, reducing computational costs and enabling more experimentation. We developed emulators to predict annual−mean fine particulate matter (PM2.5 ) and ozone (O3 ) concentrations and their associated chronic health impacts from changes in five major emission sectors (residential, industrial, land transport, agriculture, and power generation) in China. The emulators predicted 99.9% of the variance in PM2.5 and O3 concentrations. We used these emulators to estimate how emission reductions can attain air quality targets. In 2015, we estimate that PM2.5 exposure was 47.4 μg m −3 and O3 exposure was 43.8 ppb, associated with 2, 189, 700 (95% uncertainty interval, 95UI: 1, 948, 000–2, 427, 300) premature deaths per year, primarily from PM2.5 exposure (98%). PM2.5 exposure and the associated disease burden were most sensitive to industry and residential emissions. We explore the sensitivity of exposure and health to different combinations of emission reductions. The National Air Quality Target (35 μg m −3 ) for PM2.5 concentrations can be attained nationally with emission reductions of 72% in industrial, 57% in residential, 36% in land transport, 35% in agricultural, and 33% in power generation emissions. We show that complete removal of emissions from these five sectors does not enable the attainment of the WHO Annual Guideline (5 μg m −3 ) due to remaining air pollution from other sources. Our work provides the first assessment of how air pollution exposure and disease burden in China varies as emissions change across these five sectors and highlights the value of emulators in air quality research. Plain Language Summary: The ability of air quality models to help address important public health problems is limited by their high computational costs. Machine learning models can help by accurately representing these complicated air quality models for specific prediction tasks. These machine learning models can then be run many times at a fraction of the time and cost. Here, we developed machine learning models to predict long–term air quality and the associated health impacts in China from changes in emissions. We found that reducing emissions linearly improves particulate air quality and public health. The fractional improvements in public health were smaller than the fractional improvements in air quality. Removing emissions from five key sectors (residential, industrial, land transport, agriculture, and power generation) does not attain the World Health Organization Annual Guideline because of remaining pollution from other sources, such as from alternative anthropogenic emissions inside China, anthropogenic emissions outside China, and natural emissions. This work illustrates the broad reach of emulators in air pollution research. Key Points: We developed emulators to predict air quality and chronic health impacts across China from changes in emissions (based on 2015 data) Annual−mean PM2.5 exposure and the associated disease burden were most sensitive to changes in industrial and residential emissions Removing emissions from five key sectors in China does not attain the World Health Organization Annual Guideline due to remaining pollution from other sources