FY-2G云量产品与地面观测云量对比分析

Compare and Analyze FY-2G Cloud Products to Ground-Based Manual Observed Cloud Amount

为推动地面云观测自动化,利用2015 2016年全国范围内不同时段FY-2G卫星观测云覆盖率和总云量反演产品与同时段地面气象站人工观测总云量资料进行对比分析,评估卫星观测与地面人工观测的一致性和偏差。结果表明,FY-2G卫星观测云产品较地面观测偏低,总云量较云覆盖率偏低明显。定义晴天、少云、多云、阴天四种不同云量等级,进一步分析卫星数据,结果显示不同云量等级下云覆盖率与总云量与地面人工观测的一致性和偏差有所不同,晴天和少云状态下总云量产品一致性较好,阴天时云覆盖率一致性较好。从分布上分析发现西部和西南部观测偏差较大,且根据云量等级呈现不同的状态。因此在云观测自动化布局中,卫星观测不能完全替代地面云量观测。地面观测应在西部和西南部,以及天气状况较为复杂的区域加强观测。

In order to promote the layout of automated ground-based cloud amount measurements, the cloud frac- tion （CFR） and the cloud total amounts （CTA） measured by geostationary satellite FY-2G were compared to the manual cloud amount measured at surface stations to analyze the concordance rate and deviation between each other. China was selected as study area and five months （ July and October of 2015 and January, April and July of 2016） were selected to represent spring, summer, autumn and winter. There are five times manual cloud amount measurement （8 h, 11 h, 14 h, 17 h, 20 h） per day at 838 stations in China. The satellite data within 5 km radius of these station＇ s points were used to produce point＇ s data corresponding the manual cloud amount observed data, where the inverse distance to power method was used to the CFR data and the regional average method was used to the CTA data. It is considered consistent when the difference of two data is less than 2 tenths of sky cover at the same times. The results showed that the data of the satellite FY-2G cloud products＇ are usual- ly lower than that of the manual observed at surface and the CTA is more obvious than the CFR. Then four levels of sky covers were divided by the manual cloud amount for further study. The level one is clear sky that cloud amount is less than 1 tenths of sky, the level two is partly cloudy that cloud covers 2-3 tenths of sky, the level three is cloudy that cloud covers 4-7 tenths of sky, and the level four is overcast that cloud amount is more than 8 tenths of sky. The comparison between satellite data and surface observed data was conducted time by time in each level. The results indicated that the performance of CTA and CFR are not good enough in all levels of sky cover and all over the country. In average, CTA is better than CFR in the condition of clear sky and overcast sky, while CFR is better than CTA in clear sky and partly cloudy sky. For the exploration of the performance＇ spatial distribution, the west and southwest part of China suffer from scarce capacity, especially in Yunnan, Guizhou, Guangxi and Guangdong province. So, we should plan more automated ground-based cloud observa- tion in west and southwest part of China. It is obvious that although the satellite cloud product＇ could help us know more spatial information of cloud, a certain density of ground-based observation is needed to correct and promote the satellite products for further applications.