针对台风风暴潮业务的多源多维数据快速提取技术  

Rapid data extraction technology for multi-source multi-dimensional data in typhoon storm surge forecasts

作　　者：刘思晗 蔡文博[1,2] 李飞[1] 王凤菊 叶文琦 LIU Sihan;CAI Wenbo;LI Fei;WANG Fengju;YE Wenqi(National Marine Environmental Forecasting Center,Beijing 100081,China;Key Laboratory of Marine Hazards Forecasting,National Marine Environmental Forecasting Center,Ministry of Natural Resources,Beijing 100081,China;International Business Machines Corporation,Beijing 100027,China)

机构地区：[1]国家海洋环境预报中心,北京100081 [2]国家海洋环境预报中心自然资源部海洋灾害预报技术重点实验室,北京100081 [3]国际商业机器(中国)有限公司,北京100027

出　　处：《海洋预报》2024年第6期53-61,共9页Marine Forecasts

基　　金：国家自然科学基金(42076214)。

摘　　要：分析当前海洋大数据处理及提取存在的信息化问题,以台风风暴潮预报系统为例阐述了多源异构海洋环境数据的持久化保存以及快速提取关键技术。针对台风风暴潮预报业务数据读取服务场景进行深入分析,并针对该业务的海洋环境数据存储和提取提出了基于OldSQL+NoSQL+分布式文件系统的解决方案。该方案经过实际场景验证,可以有效解决台风风暴潮数值预报业务化系统中海量多源异构数据的存储和提取问题,显著提升数据提取性能,可扩展性较强。This paper analyzes the informatization management in the marine big data age,and describes the key technologies for persistent storage and rapid extraction of multi-source heterogeneous marine environmental data in typhoon storm surge forecasts.This study proposes an OldSQL+NoSQL+distributed file system based solution for marine environmental data storage and extraction in typhoon storm surge forecasts.The solution has been verified in actual scenarios and has been proven to be efficient and scalable in solving the massive multi-source heterogeneous data storage and extraction in the typhoon storm surge forecasts.

关 键 词：台风风暴潮 数据管理 海洋环境数据 关系型数据库 

分 类 号：P731.23[天文地球—海洋科学]