气象数据在多种类数据库上查询统计能力初探  

Preliminary Study on the Query and Statistics Abilityof Meteorological Data on Various Databases

作　　者：廖婷婷 支亚京 李进讷 虞雪莹 汪华[1] LIAO Tingting;ZHI Yajing;LI Jinne;YU Xueying;WANG Hua(Guizhou Meteorological Information Center,Guiyang 550002,China;Meteorological Bureau of Wudang District,Guiyang City,Guizhou Province,Guiyang 550018,China)

机构地区：[1]贵州省气象信息中心,贵州贵阳550002 [2]贵州省贵阳市乌当区气象局,贵州贵阳550018

出　　处：《中低纬山地气象》2023年第3期108-112,共5页Mid-low Latitude Mountain Meteorology

基　　金：贵州省科技支撑计划项目([2017]2819号):基于大数据分布式技术的信息数据管理技术研究;贵州气象局重要业务科研项目(黔气标合ZY[2020]09号):基于大数据的多源气象数据存储共享及支撑技术研究。

摘　　要：气象结构化数据包含地面小时数据、地面分钟数据等多种数据种类,在分析和调用中需要利用时间属性和空间属性。随着数据量日益增大,高并发、复杂统计条件下的查询与统计对数据库的效率要求十分严苛。在此背景下,传统的关系型数据库逐渐难以满足实时应用需求。该文使用不同类型的分布式数据库和关系型数据库,结合气象业务使用场景,探索多种类型下数据库的使用性能,研究不同场景的性能差异与不同数据库架构框架之间的关系,以提高数据服务实时响应能力。The meteorological structured data includes a variety of data types,such as ground hourly data,ground minute data,etc.Time attributes and spatial attributes need to be used in the analysis and call.With the increasing amount of data,queries and statistics under the conditions of high concurrency and complex statistics have strict requirements on the efficiency of the database.In this context,the traditional relational database is gradually difficult to meet the needs of real-time applications.The paper uses different types of distributed databases and relational databases,combined with meteorological business use scenarios,to explore the performance of databases under various types,and study the relationship between the performance differences of different scenarios and different database architecture frameworks to improve the real-time response capability of data services.

关 键 词：分布式数据库 关系型数据库 气象数据 

分 类 号：TP311.52[自动化与计算机技术—计算机软件与理论]