An improved meteorological variables-based aerosol optical depth estimation method by combining a physical mechanism...

Area of Science:

• Atmospheric Science
• Remote Sensing
• Geospatial Analysis

Background:

• Aerosol optical depth (AOD) retrieval using meteorological variables can be limited by data gaps and accuracy.
• Existing methods like the Elterman retrieval model (ERM) require enhancement for improved spatiotemporal coverage and precision.

Purpose of the Study:

• To develop and evaluate an improved AOD retrieval method by integrating spatiotemporal linear mixed-effect (STLME) and geographically weighted regression (GWR) models.
• To enhance the accuracy and data coverage of AOD estimation over the Beijing-Tianjin-Hebei (BTH) region.

Main Methods:

• A two-stage model, STG-ERM, was developed by combining STLME and GWR models.
• The STG-ERM model was applied to meteorological data from the BTH region for 2019 and 2020.
• Retrieval results were cross-validated against Multi-Angle Implementation of Atmospheric Correction (MAIAC) AOD data.

Main Results:

• The STG-ERM model significantly increased data coverage by 39.0% in 2019 and 40.5% in 2020.
• Cross-validation showed substantial improvements over previous models, with high determination coefficients (R²=0.86) and acceptable prediction errors.
• Fused annual mean AOD revealed distinct spatial variations, with higher values in plains and lower values in mountainous areas, and strong seasonal patterns.

Conclusions:

• The STG-ERM model provides a robust and accurate method for AOD retrieval, outperforming earlier meteorological models.
• Meteorological data coverage impacts fused AOD accuracy, with greater sensitivity in areas of high AOD.
• Continuous, high-resolution meteorological data can further enhance model performance, making STG-ERM valuable for filling gaps in satellite AOD products.