| 引用本文: | 吴奕霄， 王珂清，程 婷， 项 瑛， 周学东， 徐 芬. 一种基于阵风锋时空变化特征的大风识别算法研究[J]. 雷达科学与技术, 2023, 21(5): 547-554.[点击复制] |
| WU Yixiao, WANG Keqing, CHENG Ting, XIANG Ying, ZHOU Xuedong, XU Fen. Research on a Gale Identification Algorithm Based on Temporal and Spatial Variation Characteristics of Gust Front[J]. Radar Science and Technology, 2023, 21(5): 547-554.[点击复制] |
|
| 摘要: |
| 本文根据易引发大风的阵风锋在天气雷达反射率因子图像空间上呈现高度较低、形态呈弧线结构和时间上具有移动特点这3个特征，在低仰角反射率因子、时间差分反射率因子、空间差分反射率因子这3种不同反射率因子图像上，采用模板匹配技术分别对其进行阵风锋图像特征匹配分值处理。融合3种特征匹配结果后，采用区域增长法提取阵风锋弧线图像结构，排除虚假图像。在处理3种得分值结果的过程中，同时采用了风暴抑制模板和放射状回波抑制模板技术来降低非阵风锋回波得分值。结果表明，算法能够有效识别易引发大风的阵风锋图像特征，且识别结果与原阵风锋位置和形状较为一致。本文所设计的阵风锋模板匹配技术为能准确地监测和预警阵风锋灾害性天气过程提供多一种可能。 |
| 关键词: 模板匹配技术 区域增长法 阵风锋 特征融合 |
| DOI：DOI:10.3969/j.issn.1672-2337.2023.05.011 |
| 分类号:P412.25；TN959.4 |
| 基金项目:国家自然科学基金气象联合基金项目（No.U2142203）；中国气象科学研究院基本科研业务费专项资金项目（No.2021Z003） |
|
| Research on a Gale Identification Algorithm Based on Temporal and Spatial Variation Characteristics of Gust Front |
|
WU Yixiao, WANG Keqing, CHENG Ting, XIANG Ying, ZHOU Xuedong, XU Fen
|
|
1. Jiangsu Climate Center, Nanjing 210041, China;2. Jiangsu Weather Modification Center, Nanjing 210041, China;3. Jiangsu Meteorological Information Center, Nanjing 210041, China;4. Nanjing Joint Institute for Atmospheric Sciences, Nanjing 210041, China
|
| Abstract: |
| The gust front, which is easy to cause gale, has three features in the weather radar reflectivity factor image?low height in space, arc structure and moving characteristics in time. The template matching technology is used to process the gust front image feature matching score separately on three different reflectivity factor images: low?elevation reflectivity factor, time?difference reflectivity factor and spatial?difference reflectivity factor. After three feature matching results being fused, the region growing method is used to extract the arc?shaped structure of the gust front images and exclude false images. In the process of obtaining three kinds of score values, the storm suppression template and radial echo suppression template are also used to reduce the scores of non?gust front echo. The results show that the algorithm can effectively recognize the image features of gust front which can easily cause gale, and the identification results are consistent with the original gust fronts in terms of positions and shapes. The gust front template matching technique developed in this paper provides another possibility for accurately monitoring and warning gust front disaster weather events. |
| Key words: template matching technique region growing method gust front feature fusion |
