The Remote Sensing Retrieval Of PM_2.5 Based On MODIS Data

In recent years,as the urbanization continues to accelerate and human activity intensifies,severe fog and haze weather frequently occurs,seriously affecting human traffic travel and health.As the primary pollutant of fog and haze pollution,PM2.5 plays an negaive role in the near surface environment and human health.However,PM2.5 ground monitoring network in China was built late and the number of monitoring sites is not enough,unable to monitoring PM2.5 in large areas.It is an urgent need of high temporal and spatial resolution of PM2.5 data in big cities with large population and complex land cover types.The rapid progress of satellite remote sensing technology compensating the disadvantages of ground monitoring,which obtains a wide range,long time sequence PM2.5 concentration data,provides convenience and possibility for the assessment of the spatial and temporal distribution and health effects of air pollution.Therefore,this study takes Beijing-Tianjin-Hebei region,which is the most typical air pollution area,as the research area.Based on the MODIS image data,the PM2.5 concentration data of Beijing-Tianjin-Hebei region in December 2015 is retrieved,and the main conclusions are as follows:1)A cloud detection algorithm is improved,making the full use of the multispectral advantages of MODIS data to separate snow and ice,using visible bands to extract thick cloud and using the combination of MODIS band 8 and 26 to extract thin cloud and cirrus cloud.The snow and cloud information is detected perfectly.It is suitable for urban areas with complex surface types and atmospheric condition.2)An aerosol retrieval algorithm based on SARA(A Simple high resolution MODIS Aerosol Retrieval Algorithm)is improved,with the minimum reflectivity technology calculating surface reflectance.The retrieval AOD data with 250 m spatial resolution,has high consistency with the AERONET aerosol ground observation data,the total coefficient of fit determination reaching 0.9.Based on the retrieved AOD data,the complete PM2.5 data in Beijing-Tianjin-Hebei region is obtained by the three stage GAM model inversion,effectively expanding the extent of monitoring area.3)Based on the retrieved PM2.5 data,the spatial distribution pattern and potential health effects of PM2.5 in Beijing-Tianjin-Hebei region are analyzed and the conclusions are as follows: artificial surface is the area with the most intense human activities where air pollution is most serious,and also has a negative impact on the surrounding area;the air pollution of Beijing Tianjin and Hebei region in December 2015 is heavy,an average of more than 83.19% of the area and more than 97.46% of the population is affected by air pollution.