摘要
使用2006年8月的31次FY-2C卫星资料进行地理定位研究,发现定位误差小于0.13条扫描线,此精度满足云分类研究需求;在此基础上,使用云分析法和聚类分析法对云进行分类试验,分类结果与FY-2C业务云产品进行对比。结果表明,云分析法确定的云型比业务云产品类型多,并对中低云有很好的识别;而聚类分析结果要优于业务云产品分类结果,在云分布细节上接近实际情况。
The tests of navigation are carried out using 31 FY-2C images in Aug. 2006, The result shows that the navigation error is less than 0.13 scan line, which is good enough for cloud classification analysis. Two methods of cloud classification (nephanalysis and cluster analysis) are presented and the classifying distributions are compared with the FY-2C NOM cloud product. Cloud categories obtained in nephanalysis are more than FY-2C NOM cloud product, and middle/low clouds are recognized well by nephanalysis. The results by cluster analysis are better than NOM cloud product, and cloud distribution are more realistic.
出处
《北京大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2009年第2期257-263,共7页
Acta Scientiarum Naturalium Universitatis Pekinensis
基金
高等学校博士学科点专项科研基金(20070001003)
国家科技支撑计划项目(2006BAC12B02)资助