As the amount of environmental data expands exponentially worldwide, researchers are challenged to efficiently analyze data maintained in multiple data centers. Because distributed data access, server-side analysis, multinode collaboration, and extensible analytic functions are still research gaps in this field, this paper introduces a collaborative analysis framework for gridded environmental data, i.e. CAFE. Multiple CAFE nodes can collaborate to perform complex data analysis. Analytic functions are performed near where data are stored. A web-based user interface allows researchers to search for data of interest, submit analytic tasks, check the status of tasks, visualize the analysis results, and download the resulting data files. CAFE facilitates overall research efficiency by dramatically lowering the amount of data that must be transmitted from data centers to researchers for analysis. The results of this study may lead to the further development of collaborative computing paradigm for environmental data analysis.

随着全球环境数据量呈指数级增长，研究人员面临着有效分析多个数据中心中维护的数据的挑战。由于分布式数据访问，服务器端分析，多节点协作和可扩展的分析功能仍是该领域的研究空白，因此本文介绍了网格化环境数据的协作分析框架，即CAFE。多个CAFE节点可以协作执行复杂的数据分析。在存储数据的位置附近执行分析功能。基于Web的用户界面允许研究人员搜索感兴趣的数据，提交分析任务，检查任务的状态，可视化分析结果以及下载结果数据文件。CAFE大大降低了必须从数据中心传输到研究人员进行分析的数据量，从而提高了整体研究效率。这项研究的结果可能会导致用于环境数据分析的协作计算范式的进一步发展。