基于高光谱遥感影像的AOD反演

AOD inversion based on hyperspectral remote sensing image

气溶胶光学厚度(Aerial Optical Depth,AOD)是重要的大气参数,也是大气校正中不可缺少的变量。文中通过两种AOD反演算法的实施和比较,探讨了利用高光谱进行AOD反演的可行性和算法的适用性。通过两种方法,对四幅影像进行了气溶胶的反演,并以地基观测数据进行了对比验证。结果显示改进的V5.2算法在非浓密植被区域的精度较高,而在浓密植被区域与DDV算法精度相当,由于两种算法均依赖于波段固定的线性关系,使得两种算法在气溶胶的反演中均有误差。

Aerosol Optical Depth （AOD） was the important atmospheric parameter and the indispensable variable in the atmospheric correction. In the paper, through the implementation and comparison of two AOD inversion algo- rithms, the feasibility of the inversion and the applicability of the algorithm using hyperspectral AOD were discussed. The aerosol inversion of four images was carried out by two methods, and the ground observation data were used to verify. The results showed that the improved VS. 2 algorithm had high accuracy in the non - dense vegetation area, equaled the DDV algorithm accuracy in dense vegetation area. Since both algorithms relied on the band fixed linear relationship, two algorithms had errors in the aerosol inversion