Aerosol Optical Depth Inversion Of Beijing With HJ Remote Sensor Data

Aerosol becomes a major pollutant which affects the air quality. Small diameteraerosol in the atmosphere can also be inhaled resulting in hazards to human health.Therefore, AOD as the main parameters of aerosol optical depth has become animportant field of air pollution research. HJ satellite as environment and disastermonitoring satellite, with high spatial resolution of30meters, has a strong advantagecompared to the traditional MODIS aerosol pollution monitoring data. In this paper,satellite data for HJ surface reflectance band made amendments on the basis of a fullanalysis of the characteristics of HJ satellite data and the establishment ofgeographical features lookup for Beijing aerosol was formed. Finally we got aerosoloptical thickness distribution and PM10concentration of Beijing in January2014.The results achieved in the papers follow:(1) This paper analyzes the advantages and disadvantages of traditional aerosoloptical depth inversion algorithm (dark dense vegetation algorithm, deep bluealgorithm). we chose deep blue algorithm and made improvements to getreasonable results aerosol optical thickness based on Beijing winter geographicalfeatures in January and the band features of HJ satellite data. We realize thealgorithm by ENVI5.0and IDL.(2) The results are validated by ground-based measurement of Beijing station in theworldwide Aerosol Robotic Network(AERONET) and MODIS aerosol product.The validation and discussions show that, the AODs by deep blue algorithm hasgreat correction with aerosol data of AERONET. The results are mostly within theerror line in accuracy. The accuracy is poor when AOD value is blow0.2. Theaccuracy is good when AOD value is higher than0.2,with the relative error is21.7%.(3) The AODs over Beijing area on January in2014are used in the monitoring ofatmosphere pollution. The PM10value is calculated by AOD-PM10model. The PM10result is validated by ground observation data which is download fromBeijing Municipal Environment Center. The validation show that, the error ofAOD-PM10model reduces8.18percent from the original26%to17.82%by thehigh resolution AODs. Meanwhile the high-resolution HJ-1satellite data is betterin the display of aerosol pollution.