A method for retrieving water-leaving radiance from Landsat TM image in Taihu Lake, East China
A method for retrieving water-leaving radiance from Landsat TM image in Taihu Lake, East China
复制标题
太湖Landsat TM影像出水辐射率反演方法
DOI：
10.1007/s11769-007-0364-7
复制
发表时间：
2007
期刊：
影响因子：
--
通讯作者：
Xuezhi Feng
中科院分区：
文献类型：
--
作者：
Ronghua Ma;Guoding Kang;Deyu Wang;Xuezhi Feng
陆地卫星专题成像仪的可见和红外波段可用于内陆水研究。研究了一种利用TM图像反演太湖水体离水辐射的方法。为了估计离水辐射，在485nm、560nm和660nm三个可见光波段进行大气校正。精确计算了瑞利散射，并采用清水像元法估算了气溶胶的贡献。利用Landsat TM中红外波段（2.1 μm）识别清水沃茨，并利用野外实测数据估算了绿色波段清水像元的离水辐亮度。气溶胶散射在绿色波段的六个点，插值匹配TM图像。假设大气校正系数为1.0，得到了蓝、红波段气溶胶散射图像。基于一个简化的大气辐射传输模型，反演了三个可见波段的离水辐射。离水辐射率进行归一化，使其与其他遥感数据在不同的时间，在不同的大气条件下获得的估计。此外，还计算了水体的遥感反射率。为评价本文提出的大气校正方法，在反演模型的基础上，对校正后的遥感数据与实测水体参数进行了相关性分析。结果表明，与基于标准大气和气溶胶模型的6S（Second Simulation of the Satellite Signal in the Solar Spectrum）程序相比，基于图像本身的大气校正方法更适合于Landsat TM数据的水体参数反演。
The visible and infrared bands of Landsat Thematic Mapper (TM) can be used for inland water studies. A method of retrieving water-leaving radiance from TM image over Taihu Lake in Jiangsu Province of China was investigated in this article. To estimate water-leaving radiance, atmospheric correction was performed in three visible bands of 485nm, 560nm and 660nm. Rayleigh scattering was computed precisely, and the aerosol contribution was estimated by adopting the clear-water-pixels approach. The clear waters were identified by using the Landsat TM middle-infrared band (2.1 μm), and the water-leaving radiance of clear water pixels in the green band was estimated by using field data. Aerosol scattering at green band was derived for six points, and interpolated to match the TM image. Assuming the atmospheric correction coefficient was 1.0, the aerosol scattering image at blue and red bands were derived. Based on a simplified atmospheric radiation transfer model, the water-leaving radiance for three visible bands was retrieved. The water-leaving radiance was normalized to make it comparable with that estimated from other remotely sensed data acquired at different times, and under different atmospheric conditions. Additionally, remotely sensed reflectance of water was computed. To evaluate the atmospheric correction method presented in this article, the correlation was analyzed between the corrected remotely sensed data and the measured water parameters based on the retrieval model. The results show that the atmospheric correction method based on the image itself is more effective for the retrieval of water parameters from Landsat TM data than 6S (Second Simulation of the Satellite Signal in the Solar Spectrum) code based on standard atmospheric and aerosol models.