The Correlation Model Study Between Aerosol Optical Depth And PM2.5 Based On The Satellite Remote Sensing

In recent years, with the rapid development of China’s society and economy, aerosol is gradually becoming an important pollutant affecting the urban air quality. Fine particulate matter(PM2.5) and other atmospheric pollution problems provoke more and more people’s attention. Therefore, the use of satellite remote sensing technology to carry out air quality monitoring work has practical significance. The paper takes Beijing as the study area, develops the research of AOD retrieval with high resolution HJ-1CCD image on the basis of aerosol optical depth(AOD) inverted by medium resolution MODIS data. And analyze the relationship between AOD and PM2.5 concentration combined with the ground environment, weather station data to establish the regression fitting model to estimate the concentration of the particles, realizing the purpose of real-time dynamic monitoring of atmospheric environment.The main research results in the paper are as follows:(1) Using the improved dark pixel algorithm, the mean AOD of Beijing city was obtained in 2015 and its spatial and temporal distribution characteristics were investigated. MODIS and HJ-1 data AOD inversion accuracy are 0.838 and 0.702 respectively, and the inversion accuracy that in summer and autumn is higher than that in spring and winter. In seasonal variation, the maximum mean AOD is in summer, followed by spring, winter, autumn, and AOD in spring and summer are significantly higher than that in autumn and winter; in spatial distribution, urban and suburban AOD different greatly, with the trend of gradually increasing from northwest to Southeast.(2) The direct correlation model of MODIS and HJ-1 data AOD and PM2.5 concentration and the variation characteristics of PM2.5 in time and space were compared and analyzed. The best fitting equation between the two is quadratic function model, and the seasonal fitting model is better than the model of the year; estimation of PM2.5 concentration with HJ-1 satellite isbetter than MODIS data. The sensonal average concentration of PM2.5 in autumn and winter are higher than that in spring and summer, the four seasons concentration order is winter > autumn > spring > summer; PM2.5 concentrations spatially show gradient distribution increasing from North to South.(3) The temperature, relative humidity, barometric pressure, wind speed and other leading meteorological factors were introduced to establish the multiple regression model between HJ-1 AOD and PM2.5 concentrations.The PM2.5 concentration estimation accuracy of the fitting model is improved, resulting in a more accurate monitoring of urban air quality status.