干旱区研究 ›› 2019, Vol. 36 ›› Issue (3): 664-669.doi: 10.13866/j.azr.2019.03.17

• 天气与气候 • 上一篇    下一篇

基于图形学的致旱天气系统自动识别技术

邵建1,2, 胡文东3, 杨有林1, 裴晓蓉1, 郑鹏徽2   

  1. 1.中国气象局旱区特色农业气象重点实验室,宁夏 银川 750002;
    2.宁夏气象台,宁夏 银川 750002;
    3.成都信息工程大学大气科学学院,四川 成都 610225
  • 收稿日期:2018-07-12 修回日期:2018-11-01 发布日期:2025-10-18
  • 通讯作者: 胡文东. E-mail: hu.wendong@163.com
  • 作者简介:邵建(1981-),男,理学硕士,副研级高级工程师,从事强对流灾害预警、天气预报及数值预报释用技术研发. E-mail:shaosdh@163.com
  • 基金资助:
    宁夏回族自治区重点研发项目(2018BEG03002);宁夏青年拔尖人才工程(2017)和四川省科技厅基础应用研究重点项目(2018JY0056)共同资助

Automatic Recognition Technology of Weather Systems Resulting in Drought Based on Graphics

SHAO Jian1,2, HU Wen-dong3, YANG You-lin1, PEI Xiao-rong1, ZHENG Peng-hui2   

  1. 1. Key Laboratory of Characteristic Agrometeorological Disaster Monitoring and Early Warning and Risk Management in Arid Regions,Chine Meteorological Administration,Yinchuan 750002,Ningxia,China;
    2. Ningxia Meteorological Observatory,Yinchuan 750002,Ningxia,China;
    3. School of Atmospheric Sciences,Chengdu University of Information Technology,Chengdu 610225,Sichuan,China
  • Received:2018-07-12 Revised:2018-11-01 Online:2025-10-18

摘要: 基于图形学分析技术,采用追踪算法、矢量分析法,利用常规MICAPS数据、NECP逐日再分析资料和EC细网格资料,对我国西北地区致旱天气系统进行识别试验。通过试验,得到如下结论:① 依据天气学定义,利用追踪算法能够成功识别高空槽、高空脊等主要系统,并可得到其曲率特征;② 利用该方法识别天气尺度系统,系统中心识别率高达90%以上、平均误差小于0.5°,而高空槽识别准确率为74%~87%,平均识别误差为0.7°~1.4°。利用风场进行订正后,准确度明显提高、误差有所减小,基本可以满足干旱预测、预报业务需求。

关键词: 天气系统, 图形化, 自动识别, 追踪算法

Abstract: Based on the pattern recognition technique and weather systems which could cause drought were tested with tracking algorithm and vector analysis methods by using MICAPS,NCEP and EC-Thin data.The results showed that:① Synoptic scale troughs and ridges could be recognized with tracking algorithm,and their curvature characteristics could also be obtained; ② By using these methods,the recognition rates of pressure center could be as high as 90%,and the mean errors were lower than 0.5°.However,the recognition rates of troughs varied in a range of 74%-87%,and the mean errors in a range of 0.7°-1.4°.With revising the wind field,the accuracy was obviously improved,and the errors were reduced to some extent,which could basically meet the needs of predicting drought.

Key words: synoptic scale system, pattern recognition, auto-discrimination, tracking algorithm