Abstract: Atmospheric reanalysis products cover four-dimensional grid information of air temperature, pressure, humidity, wind speed, direction and other variables for a period of time, which can be widely applied to climate change, weapons and equipment design and many other fields. Focusing on formats and characteristics of the reanalysis product, the general framework of application platform for reanalysis products is designed with the classical three layer model of MVC, integrating the application components such as decoding, statistical analysis, visualization and standard IO interface with third-part software. The ASCII grid data requires users to export the data with GRIB format and combine data in different fields with stitching software in the past, which is time-consuming and cumbersome. A combination of grid data management and field data management results in efficient management of petabyte-scale reanalysis products with high expansion and low delay. This paper shows that the joint management combining field unit and point unit can meet the needs of the application of data retrieval in different time and space scale.

Key words: reanalysis products, GRIB, American Standard Code for Information Interchange（ASCII） grid data, metadata, field data, atmospheric application platform

摘要： 大气再分析产品涵盖了历史一段时期内的大气温度、压强、湿度、风速风向等变量的四维格点化信息，可以广泛应用到气候变化、武器装备设计等多个领域。针对再分析产品的数据格式及应用特点，再分析产品应用平台通用框架采用经典MVC三层模型，基于第三方软件开发和集成了解码、统计分析、可视化等应用组件和标准化的IO接口。以往ASCII格点数据需要用户导出GRIB格式数据，利用拼接软件拼接产生，耗时且十分麻烦。点面结合的混合数据管理，能实现对PB级再分析产品的高效管理，具有高扩展、低延迟等特点。场单元及点单元的点面结合管理方式可满足不同时空跨度的数据检索应用需求。

关键词: 再分析产品, GRIB, 美国信息交换标准代码（ASCII）格点数据, 元数据, 场数据, 大气应用平台