Constructing the 3D Spatial Distribution of the HCHO/NO₂ Ratio via Satellite Observation and Machine Learning Model

Area of Science:

• Atmospheric Chemistry
• Remote Sensing
• Air Quality Monitoring

Background:

• Satellite-derived formaldehyde-to-nitrogen dioxide ratio (FNR) is crucial for diagnosing ozone formation sensitivity.
• The accuracy of satellite FNR in representing surface conditions is debated.
• Understanding vertical ozone formation regimes is essential for effective air quality management.

Purpose of the Study:

• To develop and validate a novel approach for constructing the 3D spatial distribution of FNR in the lower troposphere.
• To investigate the differences between satellite column FNR and reconstructed surface FNR.
• To analyze trends in ozone formation sensitivity regimes over Shanghai.

Main Methods:

• Integration of satellite data with multi-axis differential Optical Absorption Spectroscopy (MAX-DOAS) measurements.
• Application of a machine learning model (Bagged trees) to reproduce vertical FNR profiles.
• Development of a fourth-order polynomial relationship between reconstruction factors and altitude.

Main Results:

• A significant difference (56.9% decrease) was observed between reconstructed surface FNR and satellite column FNR.
• A strong correlation (R=0.98) was found between reconstruction factors and altitude.
• Analysis of 2018-2022 summer data showed a trend towards transition and NO2-limited ozone formation regimes in Shanghai.

Conclusions:

• The developed 3D FNR approach enhances the accuracy of identifying surface ozone sensitivity from satellite observations.
• This method provides a more comprehensive understanding of vertical ozone photochemical formation.
• The findings highlight a changing ozone formation regime in Shanghai, necessitating adaptive control strategies.