Retrieval Method For Aerosol Optical Thickness By POLDER Data And Application Study

Atmospheric aerosol is not only an important role in global climate andquantitative remote sensing, but also a necessary parameter in air pollution. Polarizedreflectance can distinguish atmospheric scattering from surface reflection, the paperstudys the method retrieval of the aerosol optical thinkness (AOT) over urban areausing the multi-angle data of polarization from POLDER, an algorithm that based onthe look-up table with6S vector radiative transfer model, and combined totalreflectivity with polarized reflectivity was studied. The result is used to analyze thetemporal variation and spatial distribution of AOT, and the model between retrievedAOT and the ground monitor Air Pollution Index data is founded. The mainconclusions are as follows:1. Some of the results are compared with POLDER aerosols products andAERONET solar direct radiance measurement in Beijing, which show that thecorrection coefficient are0.9808and0.9229and the average relative errors are7.79%and6.1%, it also shows that the retrieved values are smaller than POLDER aerosolsproducts.2. AOT during2005~2010over Beijing are investigated based on POLDER dataand the result is used to analyze the temporal variation and spatial distribution of AOT.The statistic result showed that high values are observed in southeast while low valueswere in northwest. The AOT is decrease slowly year by year during2005~2010.Monthly average AOT have significant differences, reaching the maxmum0.65andthe minimum0.20in July and November. The AOT in spring and summer is higherthan that in autumn and winter, and it reached the maxmum and minimum in summerand autumn respectively.3. The paper studys the relationship between retrieved AOT and the groundmonitor Air Pollution Index data, the model is founded by distinct seasons.Comparing each model, it shows that the summer and autumn model are better than others, and the correction coefficient are0.9808and0.9229. When the threshold is25,the forecast accuracy are more than70%except spring, the forecast accuracy is66.67%in spring.Finally, the city’s air pollution levels are predicted by the retrieved AOT fromsatellite data, and the comparison results proved the feasibility of this approach.