气象数据在多种类数据库上查询统计能力初探

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。本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。

A preliminary study on the query and statistics ability of meteorological data on various databases

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. This 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.