陕西省智能网格天气预报系统云化改造关键技术研究与应用  

Research and Application of Key Technologies for Cloud Transformation of Shaanxi Intelligent Grid Weather Forecast System

作　　者：何林[1] 吉庆 霍庆 戴昌明[2] 陈婧 He Lin;Ji Qing;Huo Qing;Dai Changming;Chen Jing(Shaanxi Meteorological Information Center,Xi’an 710014,China;Shaanxi Meteorological Observatory,Xi’an 710014,China;National Meteorological Information Center,Beijing 100081,China)

机构地区：[1]陕西省气象信息中心,西安710014 [2]陕西省气象台,西安710014 [3]国家气象信息中心,北京100081

出　　处：《气象与环境科学》2025年第2期104-111,共8页Meteorological and Environmental Sciences

基　　金：中国气象局青年创新团队项目(CMA2024QN08);国家气象信息中心重点创新团队项目(NMIC-2024-ZD11);国家气象信息中心“揭榜挂帅”重点科技项目(NMICJBGS202416)。

摘　　要：为适应“云+端”新型气象业务技术体制改革的发展要求,实现核心业务应用系统融入气象大数据云平台“天擎”运行,从数据流分析、算法剥离、服务接口调用、监控对接等环节入手,提出陕西省智能网格天气预报系统“云化改造”的总体技术思路。对数据源切换、产品回存与发布、算法改造与重构、监视和告警信息发送、中间件服务和应用端改造等关键技术进行了研究与实现,并选取核心数据源时效、数据产品存储时长、算法运行效率、产品共享效果等关键指标对比改造前后的系统性能。结果表明:通过“云化改造”,智能网格预报业务实现由“天擎”云服务提供支撑,应用端更为精简和轻量化;数据流得到合理优化,数据产品避免了重复存储和搬移,集中管理、时效更高;算力、存储资源迁移到云,使用效率得到有效提升;核心算法和数据产品的规范化发布与共享;业务监控统一纳入气象综合业务实时监控系统“天镜”,减轻了繁重的运维工作。在“硬件、数据、流程、平台、监控”等气象信息系统集约化方面的效益显著提升。In order to adapt to the development demand for the structural reform of new meteorological operational technology of“cloud+client”,and achieve the integration of core application systems into the operation of the meteorological big data cloud platform“CMADAAS”,we propose the overall technical ideas of cloud transformation of Shaanxi Intelligent Grid Weather Forecast System“SIGMA”from important links such as the aspects of data flow analysis,algorithm stripping,service interface calling,and monitoring docking.Key technologies such as data source switching,product writing back and release,algorithm transformation and reconstruction,monitoring and alarm information transmission,middleware services and application side transformation are studied and implemented.At the meantime,key indicators such as core data source timeliness,data product storage duration,algorithm operation efficiency,and product sharing effect are selected to compare the system performance before and after the transformation.The results show that through cloud transformation,the implementation of intelligent grid weather forecasting operation is supported by the“CMADAAS”cloud service,the application client is slimmed and more lightweighted,the data flow is reasonably optimized,the data products avoid repeated storage and moving,and the centralized management and timeliness are higher.Computing power and storage resources are migrated into the cloud,and the use efficiency is effectively improved.Core algorithms and data products are published and shared in a standardized manner.Operational monitoring is integrated into the integrated meteorological operation real-time monitoring system“Tianjing”,which reduces the heavy operation maintenance work.The benefits of meteorological information systems in terms of intensification such as“hardware,data,processes,platforms and monitoring”have been significantly improved.

关 键 词：智能网格预报 秦智 天擎 算法 数据 

分 类 号：TP311.11[自动化与计算机技术—计算机软件与理论]