摘要
目前,基层台站在雷暴预警预报中主要依赖传统经验预报方法,而对卫星云图、大气电场、多普勒雷达回波和高分辨率数值预报产品等资料释用水平不高,存在主观定性分析多、客观定量计算少等不足。依托现有气象资料和雷暴预报方法,以雷暴预警预报系统构建为目标,从气象数据仓库构建、支持向量机雷暴预报、相似预报雷暴预报、Poor Man集合预报技术和多源数据融合分析等方面,探讨了雷暴预警预报系统构建方法,实现了雷暴中期预测、短期预报和短时临近预警三个层次的预报功能,解决了气象资料庞杂难以管理、数值预报产品释用水平不高、数值预报结果存在"跃变"、多源数据融合分析缺乏手段等问题,提高了基层台站雷暴预报工作效率和预报水平,同时也为其他地区类似系统构建提供了参考。
Currently,thunderstorm warning and forecast mainly rely on traditional empirical prediction methods in the grass-roots meteorological observatory. And it lacks of means in use of satellite images,atmospheric electric field data,Doppler radar echo data and high-resolution numerical weather prediction products. The traditional prediction methods exists the disadvantage of more subjective qualitative analysis and less objective quantitative calculation analysis. So a warning and forecast system of thunderstorm is constructed based on current weather conditions. Many methods such as meteorological data management,support vector machine,analog prediction,Poor Man ensemble forecasting method and multisource data fusion analysis and so on,are used to explore the thunderstorm warning and forecast system.And problems,for example,meteorological data management chaos,the level of numerical prediction product usage is not high,the results of numerical prediction exist ＂jump＂,multi-source data fusion analysis issue and so on,have been solved. So the efficiency of thunderstorm forecast and warning in grassroots stations is improved. At the same time,the method described in the text also provide a reference to build similar systems for other areas.
作者
李振锋
付伟基
陈红霞
刘丹军
Li Zhenfeng Fu Weiji Chen Hongxia Liu Danjun(Luoyang Electronic Equipment Test Center,Luoyang 471003.China No. 96276 Troops of PLA,Luoyang 471001 ,China Luoyang Meteorological Office, Luoyang 471000, China No. 96263 Troops of PLA, Luoyang 471003, China)
出处
《气象与环境科学》
2016年第4期126-134,共9页
Meteorological and Environmental Sciences
基金
中国气象局关键技术集成项目(CMAGJ2013M30)
河南省气象局科技计划项目(Z201401)资助
关键词
豫西
雷暴
预报
预警
支持向量机
相似预报
PoorMan集合预报
western Henan province
thunderstorm
forecasting
warning
support vector machine
analog prediction
Poor Man ensemble forecasting
