欢迎访问《环境科学学报》编辑部网站！

快速检索			高级检索

研究论文

基金项目：国家科技支撑计划项目(No.2012BAH32B03);海洋公益性行业科研专项(No.201205040);国家自然科学基金(No.41301452);广东省水资源节约和保护资金项目(No.FLXGL2014-D41)

摘要：利用ZY-3卫星数据高分辨率的特点,提出了一种基于图像自身的阴坡植被暗像元气溶胶光学厚度自动反演算法.首先,分区优选阴坡植被暗像元,基于程辐射信息估算红、蓝波段的气溶胶散射相函数、散射比.其次,在Gilabert算法基础上,增加地表漫反射项的考虑,利用简化的辐射传输方程直接解算阴坡植被及浓密植被暗像元气溶胶光学厚度.最后采用克里金插值,将多个暗像元气溶胶光学厚度推算到整景图像的分布,进而进行大气纠正.结果表明,香港地区ZY-3数据AOD反演结果与MODIS气溶胶C051产品趋势一致,ZY-3数据AOD结果在揭示繁华都市区内部的AOD差异,以及识别城市内部污染源方面更具优势.ZY-3数据大气纠正后,图像清晰度、对比度增强,统计结果显示水体及浓密植被的光谱特征与先验知识相符.

Abstract：This study developed an automated image-based AOD retrieval algorithm by using the high spatial resolution ZY-3 MUX data. Firstly, unique criteria for sub-images were used for selecting vegetation pixels under shadow, and phase function and single scattering albedo of aerosol for Blue and Red channels were estimated by path radiance. A Simplified Aerosol Retrieval Algorithm was then applied to calculate the AOD of dense vegetation and shadow vegetation pixels by adding diffusion radiance into Gilabert algorithm. Finally, Kriging interpolation was used to obtain the distribution of AOD over Hong Kong for atmospheric correction. The AOD image of ZY-3 MUX showed a good agreement with the operational MODIS aerosol product (MOD04 C051) at 10 km spatial resolution, while it can better represent the AOD distribution over urban area and more accurately identify local emission sources than MOD04 C051 standard product. The overall ZY-3 image sharpness and contrast were improved after atmospheric correction. Moreover, the statistical inverted reflectance spectra of both vegetation and water were consistent with our prior knowledge.