The empirical relationship between the PM2.5 concentration and aerosol optical depth over the background of North China from 2009 to 2011

Satellite observation has shown great promise in estimating the concentration of particulate matter at the surface. It is necessary to investigate the relationships between PM2.5 concentration and aerosol optical depth (AOD) to promote the accuracy of satellite remote sensing, especially in the high aerosol emitting region in North China. The annual concentration of PM2.5 was 33±10μg/m3 in the background of North China from 2009 to 2011, while the annual mean of AOD at 550nm was 0.23±0.10; there was a high correlation between the PM2.5 concentration and AOD in the region. The linear regression functions of the daily PM2.5 concentration (y) with the ground-observed AOD (x) and the MODIS AOD (x) were y=96.98x+6.61 (R2=0.58) and y=100.10x+12.13 (R2=0.57), respectively. The linear regression functions of the monthly PM2.5 concentration (y) with the ground-observed AOD (x) and the MODIS AOD (x) were y=91.66x+12.60 (R2=0.75) and y=79.33x+16.33 (R2=0.75), respectively. In addition, the results also demonstrated significant differences in the slopes and intercepts of the linear regression functions for different seasons. A large error resulted when any single function was used to retrieve the PM2.5 concentration using AOD. •The pollution of fine particles was very serious in North China in recent years.•AOD and PM2.5 were at least 2 and 3 times higher in the background of North China than in Europe and the U.S.•There were high correlations between PM2.5 and AOD with obvious seasonal patterns in North China.•On sunny days, AOD is valuable to retrieve the PM2.5 distribution as long as calibrated by the local PM2.5 measurement.