Mapping high resolution national daily NO₂ exposure across mainland China using an ensemble algorithm

Area of Science:

• Environmental Science
• Atmospheric Chemistry
• Air Quality Monitoring

Background:

• Nitrogen dioxide (NO₂) is a significant air pollutant impacting air quality, health, and climate.
• Accurate, high-resolution NO₂ concentration data is crucial for environmental and epidemiological research.
• Existing monitoring methods may lack the spatial coverage or resolution needed for comprehensive analysis.

Purpose of the Study:

• To develop and validate a national model for generating daily mean surface NO₂ concentrations across mainland China.
• To assess the accuracy and performance of the developed NO₂ estimation model at various scales.
• To analyze spatial distributions of NO₂, identify hotspots, and evaluate population exposure levels.

Main Methods:

• Utilized the extreme gradient boosting algorithm for model development.
• Input data included high-resolution tropospheric vertical column NO₂ densities from Sentinel-5 Precursor/TROPOMI and meteorological variables.
• Employed sample-based cross-validation to assess model accuracy, reporting R², RMSE, MPE, and MRPE.

Main Results:

• The model achieved high accuracy in estimating daily mean surface NO₂ concentrations (R²=0.83, RMSE=7.58 μg/m³).
• Demonstrated strong performance at regional and site-specific scales, outperforming or matching previous studies.
• Derived spatial distributions revealed NO₂ hotspots, with approximately 12% of the population exceeding national standards in 2019.

Conclusions:

• The developed nationwide model provides accurate, high-resolution daily surface NO₂ concentrations for China.
• The model effectively estimates monthly, seasonal, and annual NO₂ levels and population exposure.
• This tool offers a valuable resource for air quality management, epidemiological studies, and environmental research in China.