Case study of inhomogeneous cloud parameter retrieval from MODIS data
Case study of inhomogeneous cloud parameter retrieval from MODIS data
复制标题
MODIS数据非均匀云参数反演案例研究
DOI：
10.1029/2005gl022791
复制
发表时间：
2005
影响因子：
5.2
通讯作者：
B. Guillemet
中科院分区：
文献类型：
--
作者：
C. Cornet;J. Buriez;J. Riedi;H. Isaka;B. Guillemet
利用MODIS在太阳天顶角接近60°时观测到的层积云场景，进行了非均匀云和分数云的云参数检索。该方法基于多光谱、多尺度信息的神经网络技术。它允许检索六个云参数，即光学厚度和有效半径的像素平均值和标准差，分数云覆盖和云顶温度。将反演的云光学厚度和有效半径与基于均匀云假设的经典方法进行了比较。比较了250m像元观测得到的亚像元分数云量和光学厚度不均匀性；这种比较显示出相当好的一致性。云顶温度的反演也相当合适。
Cloud parameter retrieval of inhomogeneous and fractional clouds is performed for a stratocumulus scene observed by MODIS at a solar zenith angle near 60°. The method is based on the use of neural network technique with multispectral and multiscale information. It allows to retrieve six cloud parameters, i.e. pixel means and standard deviations of optical thickness and effective radius, fractional cloud cover, and cloud top temperature. Retrieved cloud optical thickness and effective radius are compared to those retrieved with a classical method based on the homogeneous cloud assumption. Subpixel fractional cloud cover and optical thickness inhomogeneity are compared with their estimates obtained from 250m pixel observations; this comparison shows a fairly good agreement. The cloud top temperature appears also retrieved quite suitably.