Model Of Remote Sensing Of Atmospheric Fine Particulate Matter (PM _2.5 ) Concentration Estimates

Haze, mainly caused by the rising fine particles (PM2.5) concentration, occurred more and more frequently in Eastern China in recent years. This situation, as well as the disastrous PM2.5, has received extensive attention from the public. Remote sensing, as one of the important technical methods in air pollution monitoring, could play a major role in the monitoring of fine particles.Taking a case study in Xianlin area, Nanjing, this paper correlated satellite aerosol optical thickness with concentration utilizing MODIS aerosol products and ground level PM2.5concentration, combining with ground measured atmospheric pollutants and meteorological factors, and setting up the annual and seasonal satellite remote sensing estimation models of PM2.5concentration. The analysis results are listed as follows.The hourly averaged PM concentration in Xianlin area shows a double peak trend while the PM2.5/PM10ratio shows a three peak pattern in a day. The highest monthly averaged PM2.5concentration happened in January, while August took the lowest. The PM2.5/PM10ratio reached its lowest point in April but got to its highest point in January and June. Seasonal averaged concentration shows winter is higher than any other season, and the PM2.5/PM10ratio reached its low in spring. The analysis of meteorological impact for PM2.5concentration shows that humidity and wind speed are the most responsible factors. Under the circumstances of high relative humidity (RH>80%), the correlation coefficients between PM2.5concentration and SO2, NO2concentration is higher than any other factors.Comparing the two methods of related humidity correction, the results show that the correlation coefficients between vertical corrected MODIS AOT and humidity corrected ground level PM2.5concentration are much higher than that between MODIS AOT and ground level PM2.5concentration. The satellite-based estimation of ground level PM2.5concentration using multiple linear regression model involving the meteorological factors makes a general improvement in correlation. The results show that satellite remote sensing should be effective in estimating the ground level PM2.5concentration.