摘要
目前各主要的数值预报中心NWP模式仅同化AIRS的晴空辐射资料,而AIRS仅10%左右视场不受云污染。为了提高资料的利用率,提出一种新的基于通道排序的云检测方法,直接寻找不受云影响的通道,考虑到仪器等一些噪声需要进行滤波,区别于目前采用的移动平均滤波,文中提出了高斯核滤波方法。结果表明,采用高斯核滤波之后的云检测不但可以快速有效地检测出不受云污染的晴空通道,且减少了有云资料的误判,从而为高光谱大气红外探测器辐射率资料在数值天气预报资料同化系统中的应用奠定基础。
The present NWP model of main numerical weather prediction center only assimilates clear sky radiation data of AIRS,while only 10% left and right fields of AIRS are free from cloud pollution.In order to improve the utilization rate of data,a new cloud detection method based on channel sorting is presented,which directly searches the cloud-proof channels,in consideration of some noises and being different from the present moving average filter,a Gaussian kernel filter method is put forward.It is shown that the method of Gaussian kernel filter could not only rapidly detect out the clear sky channel without being polluted by cloud,but also decrease the error of cloud data,therefore laying foundation for the assimilation system of hyper spectral atmospheric radiation data.
出处
《安徽农业科学》
CAS
北大核心
2011年第32期20093-20095,共3页
Journal of Anhui Agricultural Sciences
关键词
高光谱
AIRS
云检测
高斯核滤波
通道排序
Hyper-spectral
AIRS
Cloud detection
Gaussian kernel filtering
Channel sorting
