NPP VIIRS数据反演气溶胶光学厚度

Retrieval of aerosol optical depth using NPP VIIRS data

利用NPP卫星的VIIRS传感器数据,基于暗像元法反演陆地气溶胶光学厚度AOD。首先,根据红外波段的归一化植被指数NDVI来对暗像元进行识别;然后,利用6S软件进行辐射传输计算构建查找表;最后,根据VIIRS数据从查找表插值得到AOD,并对其进行海拔校正。选取华北地区作为反演实验区,获得了2013年9月1日的气溶胶分布。利用AERONET北京站太阳光度计地基观测结果对反演结果对比验证,发现二者具有显著的相关性,相关系数达到0.7920。将2013年9月1日的MODIS AOD产品与本研究反演的AOD进行比对,发现二者分布趋势一致,相关系数为0.7059,相关性显著。反演结果表明,本文算法反演陆地AOD效果较好,为大气颗粒物环境监测提供了良好方法手段和数据源。

As the next generation earth observing satellife launched by America, National Polar-orbiting Partnership （NPP） satel- lite is going to take over the Terra and Aqua satellite which is in extended service. Visible Infrared Imaging Radiometer containing Suite （VIIRS） is one of the five earth observation instrument which are taken in this satellite. The VIIRS channel for aerosol is inherited from MODIS on board the Terra and Aqua satellite. The application with the VIIRS data for the atmospheric pollution monitoring in China needs to be carried out urgently. In this paper, the terrestrial Aerosol Optical Depth （AOD） with 750 m reso- lution retrieved from the VIIRS data is studied. The Dark Dense Vegetation （DDV） method is used to retrieve the AOD in this research. First of all, the cloud pixels are checked and removed through the VIIRS cloud product. According to Normalized Difference Vegetation Index （NDVI） of the infra- red wavelength the dark pixels are identified. The Look Up Table （LUT） is built by the 6S radiative transfer model, ant then, the AOD is interpolated from the LUT. Finally, the AOD at different altitudes are obtained after the altitude correction. The North Plain of China, which includes Beijing, Tianjin, Hebei region, Shandong, Henan and Shanxi province, is selected as the area of the experiment. The result of September 1,2013 is presented. The result shows that the VIIRS data can monitor the distribution of aerosol very well. The result is validated by ground-based sun photometer measurements in the Beijing site of AERO- NET. The VIIRS AOD inversed in this study and the ground-based measurements agrees well, and the correlation coefficient is 0.7920. Also, the retrieved VIIRS AOD, which is resampling to 10 km resolution through the three convolution interpolation, is compared with the MODIS AOD product. The comparing result shows that their distribution trend is consistent, and the correlation coefficient is 0. 7059. In this study, AOD with 750 m resolution is obtained from the NPP-VIIRS data based on the DDV method. The retrieved AOD has high correlation coefficient with the ground-based measurements and the MODIS aerosol product. The result shows that the VIIRS data can monitor the terrestrial aerosol well. It will provide fine data source for the atmospheric aerosol particulates monitoring.