国家级天气预报检验分析系统建设与应用
|
摘要
为了适应精细化预报和业务管理的发展需求,国家气象中心建设开发了基于Web的国家级天气预报检验分析系统。系统分为预报检验、检验文件解析处理、检验数据查询分析与检验平台管理4个功能模块,关键技术包括标准化的数据管理、开放式的算法模块管理与调度和检验数据的可视化分析。系统建立了规范高效的检验业务数据流程,兼容处理预报分析制作系统(MICAPS)数据、GRIB2数据、城镇报数据、自动气象站数据等其他专业气象数据,涵盖了国家级省级智能网格预报、全国城镇天气预报、定量降水预报、大城市空气质量预报等数十项检验业务产品,给出了空间分布、柱状图及数据表格等展现形式。系统为全国各级预报员、模式开发人员和管理人员提供预报检验反馈信息,为各省以及国家级预报业务考核提供了信息支撑;同时系统提供逐旬、月、年度的智能网格预报以及城镇天气预报的检验结果对比,有力支撑了智能网格预报产品业务研发和业务试验。
National Verification System in CMA provides a unified verification data environment to realize unified management and service of the observation, forecast and verification data. A standardized and efficient verification operational process is established, which is compatible with multiple data including MICAPS data,GRIB2 data, NWFD data, automatic weather station data and other meteorological data. It works with dozens of verification operation such as national and provincial intelligent grid forecasting, urban weather forecast, quantitative precipitation forecast, and the air quality forecast in big cities. The verification products are displayed in spatial distribution map, histogram and data table.Verification results can provide management departments with assessment and evaluation of forecasters from different departments, and support optimizing the management and allocation of resources. On the other hand,forecasters can also examine the verification results to improve future forecasting. Moreover, these results can also indicate forecasting capacity of different models for model developers. Regular annual and monthly verification reports issued by official departments, temporary verification reports required for the assessment and evaluation of forecasters are provided by the inspection system.National Verification System is overall standardized and systematic. The construction of the system emphasizes unification of norms and interfaces, so as to standardize basic functions, operational processes,data models and data coding information standards of the system construction and enhance the expansibility of the system. The system is deployed on three Linux servers, namely Web server, database server and data processing server. By updating and upgrading the system, the efficiency of statistical query results is improved, the interactive response of the inspection system is faster, and the operational process is complete and more standardized.The system consists of 4 functional modules: Forecast verification, analysis of verification documents,query and analysis of verification data and management of verification platform. The system organizes and manages all kinds of data effectively, dispatches the verification algorithm uniformly, and is compatible with new verification methods in the future. Key technologies include standardized data management, open algorithm module management and scheduling,and visual analysis of verification data.In order to provide references for provincial development of relevant verification systems, the specific verification methods of each module and the detailed processing in real-time operation are also described in particular. At the same time,the system provides comparison between results of urban weather forecasting and ten-day, monthly and annual intelligent grid forecasting. And therefore, it strongly supports the operational research and development of intelligent grid forecasting products and operational tests.
National Verification System in CMA provides a unified verification data environment to realize unified management and service of the observation, forecast and verification data. A standardized and efficient verification operational process is established, which is compatible with multiple data including MICAPS data,GRIB2 data, NWFD data, automatic weather station data and other meteorological data. It works with dozens of verification operation such as national and provincial intelligent grid forecasting, urban weather forecast, quantitative precipitation forecast, and the air quality forecast in big cities. The verification products are displayed in spatial distribution map, histogram and data table.Verification results can provide management departments with assessment and evaluation of forecasters from different departments, and support optimizing the management and allocation of resources. On the other hand,forecasters can also examine the verification results to improve future forecasting. Moreover, these results can also indicate forecasting capacity of different models for model developers. Regular annual and monthly verification reports issued by official departments, temporary verification reports required for the assessment and evaluation of forecasters are provided by the inspection system.National Verification System is overall standardized and systematic. The construction of the system emphasizes unification of norms and interfaces, so as to standardize basic functions, operational processes,data models and data coding information standards of the system construction and enhance the expansibility of the system. The system is deployed on three Linux servers, namely Web server, database server and data processing server. By updating and upgrading the system, the efficiency of statistical query results is improved, the interactive response of the inspection system is faster, and the operational process is complete and more standardized.The system consists of 4 functional modules: Forecast verification, analysis of verification documents,query and analysis of verification data and management of verification platform. The system organizes and manages all kinds of data effectively, dispatches the verification algorithm uniformly, and is compatible with new verification methods in the future. Key technologies include standardized data management, open algorithm module management and scheduling,and visual analysis of verification data.In order to provide references for provincial development of relevant verification systems, the specific verification methods of each module and the detailed processing in real-time operation are also described in particular. At the same time,the system provides comparison between results of urban weather forecasting and ten-day, monthly and annual intelligent grid forecasting. And therefore, it strongly supports the operational research and development of intelligent grid forecasting products and operational tests.
引文
[1] Finley J P. Tornado predictions. Amer Meteor J,1884,1:85-88.
[2] Stanski II R, Wilson L J, Burrows W R. Survey of Common Verification Methods in Meteorology//World Weather Watch Technical Report No. 8. Geneva:World Meteorological Organization,1989.
[3] Jolliffe I T, Stephenson D B. Forecast Verification:A Practitioner's Guide in Atmospheric Science(Second Edition). New York:Wiley and Sons Ltd,2011.
[4] Nurmi P. Operational Verification System(s)at FMI. 6th International Verification Methods Workshop,2014.
[5] Yadav B P,Kumar N,Rathore L S. Operational Weather Forecast Verification at India Meteorological Department. 6th International Verification Methods Workshop,2014.
[6]牟惟丰.中央气象台一年来预报评分结果分析.气象,1988,14(11):49-51.
[7]唐文苑,周庆亮,刘鑫华,等.国家级强对流天气分类预报检验分析.气象,2017,43(1):67-76.
[8]束家鑫,王志烈.我国台风研究的十年进展//台风会议文集(1981).上海:上海科学技术出版社,1983:1-4.
[9]董克勤,杨麟美,周江兴.台风路径预报现状分析.气象,1986,12(7):2-6.
[10]许映龙,张玲,高栓柱.我国台风预报业务的现状及思考.气象,2010,36(7):43-49.
[11]李佰平,戴建华,张欣,等.三类强对流天气临近预报的模糊检验试验与对比.气象,2016,42(2):129-143.
[12]刘还珠,赵声蓉,陆志善,等.国家气象中心气象要素的客观预报-MOS系统.应用气象学报,2004,15(2):181-191.
[13]邓国,龚建东,邓莲堂,等.国家级区域集合预报系统研发和性能检验.应用气象学报,2010,21(5):513-523.
[14]皇甫雪官.国家气象中心集合数值预报检验评价.应用气象学报,2002,13(1):29-36.
[15]段明铿,王盘兴,吴洪宝.夏季亚欧中高纬度环流的集合预报效果检验.应用气象学报,2009,20(1):56-61.
[16]毕宝贵,代刊,王毅,等.定量降水预报技术进展.应用气象学报,2016,27(5):534-549.
[17]林明智,毕宝贵,乔林.中央气象台短期降雨预报水平初步分析.应用气象学报,1995,6(4):392-399.
[18]董立清.1991年江淮暴雨的定量预报检验.应用气象学报,1993,4(3):333-340.
[19]杨元琴,王继志,侯青,等.北京夏季空气质量的气象指数预报.应用气象学报,2009,20(6):649-655.
[2] Stanski II R, Wilson L J, Burrows W R. Survey of Common Verification Methods in Meteorology//World Weather Watch Technical Report No. 8. Geneva:World Meteorological Organization,1989.
[3] Jolliffe I T, Stephenson D B. Forecast Verification:A Practitioner's Guide in Atmospheric Science(Second Edition). New York:Wiley and Sons Ltd,2011.
[4] Nurmi P. Operational Verification System(s)at FMI. 6th International Verification Methods Workshop,2014.
[5] Yadav B P,Kumar N,Rathore L S. Operational Weather Forecast Verification at India Meteorological Department. 6th International Verification Methods Workshop,2014.
[6]牟惟丰.中央气象台一年来预报评分结果分析.气象,1988,14(11):49-51.
[7]唐文苑,周庆亮,刘鑫华,等.国家级强对流天气分类预报检验分析.气象,2017,43(1):67-76.
[8]束家鑫,王志烈.我国台风研究的十年进展//台风会议文集(1981).上海:上海科学技术出版社,1983:1-4.
[9]董克勤,杨麟美,周江兴.台风路径预报现状分析.气象,1986,12(7):2-6.
[10]许映龙,张玲,高栓柱.我国台风预报业务的现状及思考.气象,2010,36(7):43-49.
[11]李佰平,戴建华,张欣,等.三类强对流天气临近预报的模糊检验试验与对比.气象,2016,42(2):129-143.
[12]刘还珠,赵声蓉,陆志善,等.国家气象中心气象要素的客观预报-MOS系统.应用气象学报,2004,15(2):181-191.
[13]邓国,龚建东,邓莲堂,等.国家级区域集合预报系统研发和性能检验.应用气象学报,2010,21(5):513-523.
[14]皇甫雪官.国家气象中心集合数值预报检验评价.应用气象学报,2002,13(1):29-36.
[15]段明铿,王盘兴,吴洪宝.夏季亚欧中高纬度环流的集合预报效果检验.应用气象学报,2009,20(1):56-61.
[16]毕宝贵,代刊,王毅,等.定量降水预报技术进展.应用气象学报,2016,27(5):534-549.
[17]林明智,毕宝贵,乔林.中央气象台短期降雨预报水平初步分析.应用气象学报,1995,6(4):392-399.
[18]董立清.1991年江淮暴雨的定量预报检验.应用气象学报,1993,4(3):333-340.
[19]杨元琴,王继志,侯青,等.北京夏季空气质量的气象指数预报.应用气象学报,2009,20(6):649-655.