Abstract: By using GMS-4 infrared brightness temperature and visible albedo data from Jan. to Feb. In 1994, methods of extracting sea ice parameters are developed and digital remote sensing pictures are obtained in Liaodong Bay. Based on the difference in physical properties between ice and water, a criterion distinguishing ice from water is set up and the threshold value of brightness is achieved. Ice thickness has been calculated according to the relationship between ice thickness and brightness by thephysical and statistical methods. The retrieval distributions of ice thickness are in good agreement with observations. The results show that a precision of ice identification and thickness retrieval is 78.25% and 65.97%, respectively.

Key words: geostationary meteorological satellite (GMS-4), sea ice, ice-water distinguishing, ice thickness

摘要： 利用1994年1、2月的GMS-4的红外亮温和可见光反照率资料，以辽东湾海冰为对象，得到了海冰参数的提取方法和冰厚遥感数字分布。根据冰水物理特性的差异，建立了冰水识别的判据，发现用亮温可以较好地识别冰水，由此得到了冰水区分的亮温阈值；根据物理和统计方法建立了冰厚与亮温的对应关系，分别对海冰冰厚分布进行了反演，得到了与实况较为一致的冰厚分布。结果表明，冰、水区分的准确率为78.25%，冰厚反演的准确率为65.97%。

关键词: GMS-4, 海冰, 冰水识别, 冰厚