Abstract
Satellite aerosol products have been widely used to retrieve ground $PM_{2.5}$ concentration because of its wide coverage and continuously spatial distribution. While more and more studies focus on the retrieval algorithm, we find that the relationship between $PM_{2.5}$ concentration and satellite AOD has not been fully discussed in China. Is satellite AOD always a good indicator for $PM_{2.5}$ in different regions and can AOD still be employed to retrieve $PM_{2.5}$ with pollution conditions changing in these years remain unclear. In this study, the relationships between $PM_{2.5}$ and AOD were investigated in 368 cities in China for a continuous period from February 2013 to December 2017, at different time and regional scales. Pearson correlation coefficients and $PM_{2.5}$/AOD ratio were used as the indicator. Firstly, we concluded the relationship of $PM_{2.5}$ and AOD in terms of spatiotemporal variations. Then the impact of meteorological factors, aerosol size and topography were discussed. Finally, a GWR retrieval experiment was conducted to find out how was the retrieval accuracy changing with the varying of $PM_{2.5}$-AOD relationship. We found that spatially the correlation is higher in Beijing-Tianjin-Hebei and Chengyu region and weaker in coastal areas such as Yangtze River Delta and Pearl River Delta. The $PM_{2.5}$/AOD ratio has obvious North-South difference with a high ratio in north China and a lower ratio in south China. Temporally, $PM_{2.5}$/AOD ratio is higher in winter and lower in summer, the correlation coefficient tends to be higher in May and September. As for interannual variations from 2013 to 2017, we detected a declining tendency on $PM_{2.5}$/AOD ratio. The accuracy of GWR retrievals were decreasing too, which may imply that AOD may not be a good indicator for $PM_{2.5}$ in the future.
Introduction
To monitor air pollution over vast areas, scientists often use satellite-derived Aerosol Optical Depth (AOD)—a measure of how much light is blocked by airborne particles in a vertical column of the atmosphere. However, AOD measures the entire atmospheric column, whereas humans breathe $PM_{2.5}$ at the ground level. The "mapping" between these two isn't a simple 1:1 ratio. It is influenced by weather, particle size, and altitude. While many researchers focus on building complex machine-learning "black box" models to predict $PM_{2.5}$ from AOD, this study takes a step back to ask: How solid is the underlying physical and statistical relationship between these two variables across different regions and times in China?
Datasets and Methodology
The study integrated several high-quality datasets spanning nearly five years (February 2013 to December 2017):
- Ground $PM_{2.5}$: Hourly data from approximately 1,500 monitoring stations across 368 cities, provided by the Ministry of Environmental Protection of China.
- Satellite AOD: MODIS (Terra) Level 2 daily AOD at 10km resolution.
- Meteorological and Aerosol Data: NCEP/NCAR Reanalysis for humidity, temperature, and wind; MERRA-2 for Planetary Boundary Layer Height (PBLH); and ground-based sun photometer data from five AERONET sites for particle size analysis.
The researchers employed two primary indicators: the Pearson Correlation Coefficient ($r$) to measure how "in sync" the variations of $PM_{2.5}$ and AOD are, and the $PM_{2.5}$/AOD ratio ($\eta$), which indicates the mass concentration of $PM_{2.5}$ per unit of aerosol optical thickness. Multi-scale analysis (city, region, month, and year) and Geographically Weighted Regression (GWR) were used to test the impact of these relationships on retrieval accuracy.
Regional and Temporal Variations
The study identifies significant spatial variance in the $PM_{2.5}$-AOD relationship. The correlation coefficient ($r$) varied wildly across cities, ranging from a negligible 0.013 to a strong 0.876, with a national mean of 0.378.
The "Basin Effect" and Topography
Topography plays a massive role. Regions like the Sichuan Basin (Chengyu) showed very high correlations ($r \approx 0.58$). Because these areas are "closed," pollutants from local industry and vehicles stay trapped near the ground, making the satellite's vertical view a much better proxy for ground conditions.
The North-South Divide and Humidity
The $PM_{2.5}$/AOD ratio is significantly higher in North China (mean $\approx$ 201 in the Beijing-Tianjin-Hebei region) compared to the South (often below 100). In southern coastal areas like the Pearl River Delta, the correlation is much weaker. High humidity causes particles to swell with water, increasing their light-blocking ability (AOD) without necessarily increasing their dry mass ($PM_{2.5}$), thus lowering the ratio.
Seasonal and Interannual Trends
The ratio peaks in winter due to low boundary layers trapping smoke and dust, and troughs in summer. Perhaps the most critical finding is the interannual decline. From 2013 to 2017, both the correlation and the ratio showed a downward trend. As China’s air quality improves and the composition of aerosols changes, the "link" between what the satellite sees and what is on the ground is stretching thin.
Conclusion
The GWR retrieval experiment confirmed that the accuracy ($R^2$) of $PM_{2.5}$ predictions dropped significantly over the five-year period, mirroring the decline in the $PM_{2.5}$-AOD correlation. This suggests that as pollution conditions change, the role of AOD in air quality monitoring needs to be "repositioned." Future models may need to rely less on AOD and more on a combination of chemical transport models and diverse meteorological inputs to maintain accuracy.
Summary updated