基于SOA架构的碳通量数据在线处理系统开发与应用

涡度相关技术是目前CO₂通量观测使用较为广泛的方法,由于环境和气象条件等问题观测数据经常有异常或缺失,因此观测数据在使用之前需要进行一系列的校正等处理工作。中国陆地生态系统通量观测研究网络 (ChinaFLUX) 结合国内外相关研究,制定了一套30 min CO₂通量数据处理流程,主要步骤包括数据质量控制、缺失数据插补及分解和统计;并开发了ChinaFLUX桌面版通量数据处理系统,解决了传统通量观测数据处理工作中存在的处理过程繁琐、方法不统一、处理不及时等问题。但该系统自身也有一些局限性,例如用户需在本地安装客户端程序,若系统更新则需相应升级,且系统运行效率依赖于用户的软硬件平台等。为了尝试解决这些问题,本文基于面向服务的体系架构(SOA,Service-Oriented Architecture)和Web Service技术,设计并开发了便于共享、即时可用的通量数据在线处理系统,实现了不依赖于操作系统和软件平台的方便、快捷的在线通量处理,用户只需一个Web浏览器即可随时随地在线处理通量数据。本研究成功地将SOA架构应用到了ChinaFLUX通量数据处理业务上,解决了桌面版通量数据处理系统存在的问题。

Development and Application of Carbon Flux Data Online Processing System Based on SOA Structure

Eddy covariance technique is a popular and reliable method to directly measure the carbon exchanges between ecosystems and the atmosphere. However, due to some inevitable operational and meteorological constraints, the observation data often have anomalies and missing values. Therefore, a series of processing work are needed to be done before explaining ecological significance from the data. Based on the state of the art of carbon flux research, Chinese Terrestrial Ecosystem Flux Research Network (ChinaFLUX) integrated quality control and gap filling, carbon flux separation, and calculation of carbon flux in different time scales into a formulated procedure and developed it into a flux data processing system, which addressed some traditional problems, such as tedious processing, inconsistent methods and lag of processing. However, the system also had some limitations because of the employment of Client/Server architecture. For example, the users have to install client application and process their data locally, which led to the operating efficiency dependent on the users' hardware and software platform. And if there are updates in the system, all users have to update their client applications accordingly. On the view of improvement, based on the Service-Oriented Architecture (SOA) and Web Services technique, we designed and developed an online flux processing system, which is readily accessible and easy to share. This online system is independent of operation system and development platform, thus the users do not have to install any software or application and can process their data at any time only with a web browser, thereby accomplished convenient and fast flux data processing. This study successfully applied SOA architecture in the ChinaFLUX data processing and addressed the limitations of the earlier desktop system.