地理科学 ›› 2016, Vol. 36 ›› Issue (10): 1581-1587.doi: 10.13249/j.cnki.sgs.2016.10.016

• • 上一篇    下一篇

静止气象卫星资料在白天海雾动态监测中的应用

邓玉娇1(), 田永杰2, 王捷纯1()   

  1. 1.广东省生态气象中心, 广东 广州 510640
    2. 华南理工大学电信学院, 广东 广州510640
  • 收稿日期:2015-12-15 修回日期:2016-02-21 出版日期:2016-10-20 发布日期:2020-09-07
  • 作者简介:

    作者简介:邓玉娇(1980-),女,湖北荆州人,博士,高级工程师,主要从事卫星遥感应用工作。E-mail:yujiao_d@163.com

  • 基金资助:
    广东省气象局科学技术研究项目(2014B08)、公益性行业(气象)科研专项经费项目(GYHY201306042)共同资助

Dynamic Detection of Daytime Sea Fog Using Geostationary Meteorological Satellite Data

Deng Yujiao1(), Tian Yongjie2, Wang Jiechun1()   

  1. 1. Guangdong Ecological Meteorology Center, Guangzhou 510640, Guangdong, China
    2. School of Electronic and Information Engineering, South China University of Technology, Guangzhou 510640, Guangdong, China
  • Received:2015-12-15 Revised:2016-02-21 Online:2016-10-20 Published:2020-09-07
  • Supported by:
    Science and Technology Research Program of Guangdong Meteorological Service(2014B08),Special Fund for Meteorologicl Research in the Public Welfare(GYHY201306042)

摘要:

利用国产静止气象卫星FY2E数据建立白天海雾监测算法,利用VIS通道反射率实现海面与云雾区分离,利用IR1通道估算云高实现中高云与低层云雾的分离,利用VIS、IR1、IR4波段构建雾判识指数初步实现海雾与低云的分离,利用平滑稳定度指数进一步实现海雾与低云的分离,最终得到的海雾监测小时产品。根据2014年1~5月份广东沿海13个海雾监站点的实测数据,对本算法所得海雾产品进行精度检验,计算得到检测率POD为92.7%,检率FAR为29.4%,总体精度为64.7%。个例分析可知,静止卫星资料因其具备较高的时间分辨率,可较好实现对海雾过程的连续、动态监测。

关键词: FY-2E, 海雾, 光谱特性, 动态阈值

Abstract:

Using Channel VIS, IR1 and IR4 of the domestic geostationary meteorological satellite FY2E/VISSR data, the multichannel method was proposed to detect the daytime sea fog. Firstly, the existing dynamic threshold method was revised in the paper in order to improve the accuracy of distinguishing sea surface from cloud and fog. The thresholds were based on the histogram statistics of the reflectance of Channel VIS, and adjusted dynamically in different regions or different seasons. Secondly, cloud height estimated from the bright temperature of Channel IR1 was used to separate middle- and high-level cloud from low-level cloud and fog. When the height was greater than 2 000 m, the object was middle- and high-level cloud. Thirdly, fog index computed from Channel IR1,IR4 and VIS was used to divide low-level cloud from fog preliminarily. If the fog index was greater than 20, the object was possibly fog. Finally, smoothness and stability index were used to tell fog from low-level cloud further, and it was greater than 0.9 for fog. Ground-based fog observation data from thirteen sites on the coast of Guangdong were used to do the verification of the FY-2E fog detection products, which were Zhanjiang, Wuchuan, Leizhou, Yangjiang, Shangchuandao, Zhuhai, Doumen, Shanwei, Lufeng, Shantou, Nan’ao, Chaoyang and Chenghai. The statistical calculation showed that the probabilities of detection(POD) was 92.7%, the false alarm ratios(FAR)was 29.4%, and the critical success index(CSI) was 64.7%. The FAR was a little high mainly because of two reasons: on one hand, the FY2E data didn’t have the penetrability for clouds, so fog covered by clouds would be misjudged to clouds; on the other hand, the recognition capability of FY-2E data with a spatial resolution of 5 km was limited so that some fog pixels couldn’t be distinguished correctly. Both of them were decided by the remote sensor of FY-2E, not decided by the fog detection algorithm. Nonetheless, the fog detection method was generally efficient and feasible. In the case study of the fog occurred on February 26, 2014, the remote sensing products showed the dynamic change of fog every hour from 9:00 AM to 16:00 PM. The sea fog almost kept unchanged between 9:00 AM and 13:00 PM, and the fog edge dissolved slowly on the junction of Bohai Gulf and the Yellow Sea between 14:00 PM and 16:00 PM. It was seen that the high time resolution data of geostationary satellite data were more effective in continuous and dynamic detection of sea fog than other satellite data.

Key words: FY-2E, sea fog, spectral characteristics, dynamic threshold

中图分类号: 

  • P405