本文已被：浏览 1554次 下载 3275次
投稿时间：2006-09-29 修订日期：2006-11-15
投稿时间：2006-09-29 修订日期：2006-11-15
中文摘要: 基于雷达数据的风暴体识别、追踪及预警方法是最早出现的临近预报技术，也是天
气雷达系统和强天气预警业务的基本组成部分。风暴体识别、追踪及预警方法可以分为三大
类：持续性预报法、交叉相关法和单体质心法，它们都属于外推预报法。其中，持续性预报
法目前已经被后两者取代。首先较详细地介绍了交叉相关法和单体质心法的研究历史和主要
算法，然后集中介绍了近几年来在外推预报基础上发展起来的一些新方法。最后，结合悉尼
奥运会期间的FDP项目，讨论了临近预报技术的检验和准确性评价，重点介绍了列联表方法
。
Abstract:The current development on storm identification, tracking and forecast
ing based on weather radar data are reviewed. Storm identification
, tracking and forecasting, as the earliest nowcasting techniques, are the basic
and essential parts of radar and severe weather warning operations. At present,
they include three aspects: cross correlation algorithm, cell centroid cracking
algorithm, persistence prediction method, which are called as extrapolation. Among t
hem, Persistence prediction method has been replaced by the former two methods.
Firstly, the history and main algorithms of cross correlation and cell centroid
method are presented in detail. Secondly, a few new methods based on the extrapo
lation are described. Finally, in combination with FDP project during Sydney Oly
mpic Games, test and evaluation on nowcasting technique are also discussed.
keywords: storm nowcasting cross correlation centroid tracking
文章编号： 中图分类号： 文献标志码：
基金项目:本文由高等学校博士学科点专项科研基金课题(20040001008)资助
| 作者 | 单位 |
| 韩雷 | 北京大学物理学院大气科学系，北京 100871 |
| 王洪庆 | 北京大学物理学院大气科学系，北京 100871 |
| 谭晓光 | 中国气象局北京城市气象研究所 |
| 林隐静 | 北京大学物理学院大气科学系，北京 100871 |
引用文本：
韩雷,王洪庆,谭晓光,林隐静,2007.基于雷达数据的风暴体识别、追踪及预警的研究进展[J].气象,33(1):3-10.
Han Lei,Wang Hongqing,Tan Xiaoguang,Lin Yinjing,2007.Review on Development of Radar based Storm Identification, Tracking and Forecasting[J].Meteor Mon,33(1):3-10.
韩雷,王洪庆,谭晓光,林隐静,2007.基于雷达数据的风暴体识别、追踪及预警的研究进展[J].气象,33(1):3-10.
Han Lei,Wang Hongqing,Tan Xiaoguang,Lin Yinjing,2007.Review on Development of Radar based Storm Identification, Tracking and Forecasting[J].Meteor Mon,33(1):3-10.