Multiply improved positive matrix factorization for source apportionment of volatile organic compounds during the COVID-19 shutdown in Tianjin, China

Yao Gu; Baoshuang Liu; Qili Dai; Yufen Zhang; Ming Zhou; Yinchang Feng; Philip K Hopke

doi:10.1016/j.envint.2021.106979

Multiply improved positive matrix factorization for source apportionment of volatile organic compounds during the COVID-19 shutdown in Tianjin, China

Environ Int. 2022 Jan:158:106979. doi: 10.1016/j.envint.2021.106979. Epub 2021 Nov 16.

Authors

Yao Gu¹, Baoshuang Liu², Qili Dai¹, Yufen Zhang¹, Ming Zhou³, Yinchang Feng¹, Philip K Hopke⁴

Affiliations

¹ State Environmental Protection Key Laboratory of Urban Ambient Air Particulate Matter Pollution Prevention and Control & Tianjin Key Laboratory of Urban Transport Emission Research, College of Environmental Science and Engineering, Nankai University, Tianjin 300350, China; CMA-NKU Cooperative Laboratory for Atmospheric Environment-Health Research, Tianjin 300350, China.
² State Environmental Protection Key Laboratory of Urban Ambient Air Particulate Matter Pollution Prevention and Control & Tianjin Key Laboratory of Urban Transport Emission Research, College of Environmental Science and Engineering, Nankai University, Tianjin 300350, China; CMA-NKU Cooperative Laboratory for Atmospheric Environment-Health Research, Tianjin 300350, China. Electronic address: lbsnankai@foxmail.com.
³ State Key Joint Laboratory of Environment Simulation and Pollution Control, College of Environmental Sciences and Engineering, Peking University, Beijing 100871, China.
⁴ Department of Public Health Sciences, University of Rochester School of Medicine and Dentistry, Rochester, NY 14642, USA; Institute for a Sustainable Environment, Clarkson University, Potsdam, NY 13699, USA.

PMID: 34991244
DOI: 10.1016/j.envint.2021.106979

Abstract

Ambient concentrations of volatile organic compounds (VOCs) vary with emission rates, meteorology, and chemistry. Conventional positive matrix factorization (PMF) loses information because of dilution variations and chemical losses. Multiply improved PMF incorporates the ventilation coefficient, and total solar radiation or oxidants to reduce the effects of dispersion and chemical loss. These methods were applied to hourly speciated VOC data from November 2019 to March 2020 including during the COVID-19 shutdown. Various comparisons were made to assess the influences of these fluctuation drivers by time of day. Dispersion normalized PMF (DN-PMF) reduced the dispersion variations. Dispersion-radiation normalized PMF (DRN-PMF) reduced the impact of chemical loss, especially at night, which was better than Dispersion-O_x normalized PMF (DON-PMF). The conditional bivariate probability function (CBPF) plots of DRN-PMF results were consist with actual source locations. The DN-PMF, DRN-PMF, and DON-PMF results were consistent between 10:00 and 15:00, suggesting dispersion was significantly more influential than photochemical reactions during these times. The DRN-PMF results indicated that the highest VOC contributors during the COVID-19 shutdown were liquefied petroleum gas (LPG) (28.8%), natural gas (25.2%), and pulverized coal boilers emissions (19.6%). Except for petrochemical-related enterprises and LPG, the contribution concentrations of all other sources decreased substantially during the COVID-19 shutdown, by 94.7%, 90.6%, and 86.8% for vehicle emissions, gasoline evaporation, and the mixed source of diesel evaporation and solvent use, respectively. Controlling the use of motor vehicles and related volatilization of diesel fuel and gasoline can be effective in controlling VOCs in the future.

Keywords: COVID-19; Improved PMF; Source apportionment; Volatile organic compounds.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Air Pollutants* / analysis
COVID-19*
China
Environmental Monitoring
Humans
SARS-CoV-2
Vehicle Emissions / analysis
Volatile Organic Compounds* / analysis

Substances

Air Pollutants
Vehicle Emissions
Volatile Organic Compounds