论文部分内容阅读
采用我国地面报资料、国际船舶报资料和云卫星(CloudSat)产品,分别对多通道动态阈值云检测方法的云检测结果进行了检验。结果表明,动态阈值云检测方法可用于长时间序列卫星观测数据的云检测,不仅运行速度较快,可实现全自动,且精度较高,陆地上空云检测准确率为82.8%(地面报资料),海洋上空云检测准确率为91.5%(船舶报资料)。此外,还利用CloudSat资料检验不同观测时间差对NOAA系列卫星云检测结果检验的影响。
The cloud detection results of multi-channel dynamic threshold cloud detection method were tested by using China’s surface information, international shipping information and CloudSat products. The results show that the dynamic threshold cloud detection method can be used for cloud detection of long time series satellite observation data, which not only runs fast but can be fully automatic with high accuracy. The accuracy of cloud detection over land is 82.8% (ground information) The accuracy of cloud detection over the ocean was 91.5% (Ship Newspaper). In addition, CloudSat data is also used to test the effect of different observation time differences on the NOAA series satellite cloud inspection results.