Abstract: Based on the difference characters of cloud, cloud classification methods are presented by using GMS-5 data, and results of cumulonimbus, cumulus congestus, cirrus, cumulus, middle cloud, dense cloud are obtained. The comparison with the ground observation data of Shouxian (Anhui Province of China) in June and July(1998) shows a satisfied results, with an accuracy of nearly 70%.

Key words: geostationary meteorological satellite(GMS-5), brightness temperature, cloud classification, multichannel

摘要： 利用GMS-5红外3个通道(2个分裂窗区通道和水汽通道)的亮温资料，依据各种云不同的辐射特征对其进行分类，由此识别出积雨云(Cb)、浓积云(Cg)、卷云(Ci)、积云(Cu)、中云(Cm)、密云(Dense)等6种主要云类以及雾(Fog)和晴空区域。结合安徽寿县地面气候基准站1998年6、7月份的地面观测资料，对分类结果进行了检验，得到了较为令人满意的结果，准确率接近70%.

关键词: GMS-5, 亮温资料, 云的分类识别, 多通道