Collaborative Research: Earth System Curator: Spanning the Gap Between Models and Datasets

“地球系统管理员”合作项目将统一处理与气候变化有关的模型和数据集，方法是开发一种通用语言——元数据形式主义——用来描述这两者，并基于这种形式主义建立一套工具原型，使研究人员能够无缝地、轻松地操作模型和数据集。最终的目标是提高气候研究人员的生产力和对地球系统的理解。合作者包括麻省理工学院、普林斯顿大学、佐治亚理工学院和国家大气研究中心的计算机科学和地球科学研究人员。提出的工作建立在两个正在进行的社区努力，地球系统建模框架（ESMF）和地球系统网格II （ESG）。ESMF是一项国家倡议，旨在为国家气候和天气模型开发通用建模基础设施，包括耦合工具和标准建模实用程序。ESG的主要目标是通过使用网格、数据/元数据和门户技术，为全国的全球变化影响研究人员提供高分辨率、长时间的气候模型模拟结果。该团队将探索ESMF和ESG可以有效结合的那些方面，并将建立一个新实体的原型，即地球系统管理员，以跨越两者之间的差距。策展人以一个至关重要的见解开始：用于全面指定模型配置的描述符也需要用于对模型输出数据进行科学有用的描述。开发描述这两者的通用元数据模式将成为这个独特而强大的社区资源的基础。策展人将提供一个社区数据库，研究人员可以从中存档和查询各种地球系统模型、实验、模型组件以及模型输出数据和结果。研究人员随后将能够分析先前运行的模型输出，或者访问模型并修改并自己运行它，无论是在本地计算机上还是在计算网格的虚拟化资源上。除了查询功能之外，该项目还将对一个工具进行原型化，该工具将测试模型组件集或数据集是否可以相互作用以形成应用程序。最后，作为策展人工作的一部分，用于自动生成组件包装器和应用程序的工具也将被原型化。策展人是在气候和相关研究中使用信息技术的社区愿景的一部分。策展人原型将有助于进一步建议和定义下一代建模和数据管理工具的形式，通过提供具体的表示来增加创新想法的可信度。此外，ESMF和ESG的共同研究人员，由于他们的项目强调生产软件和广泛的客户基础，在nsf资助的原型阶段之后，他们处于一个很好的位置，可以将策策者工具转变为可行的产品。在策展人努力的过程中，许多与ESMF和ESG相关的研究人员将被鼓励尝试并提供策展人软件的反馈。策展人项目取得的进展将通过会议和出版物，以及基于多组件高性能计算建模的共同需求、信息存档和访问的便利性以及模拟数值的相似性而建立的关系网络，影响其他领域。虽然直接归功于策展人的努力本身在气候预测方面取得的进展可能既难以追踪又温和，但地球系统研究的综合环境对于解决不久的将来的世界气候问题至关重要，而策展人是朝着这个方向迈出的决定性一步。策展人还将被纳入佐治亚理工学院软件工程课程的项目基础设施。将准备适当的介绍性材料并定义项目机会。许多项目机会将采取向公众提供气候数据的形式。通过这种方式，策展人不仅为软件工程学生提供了一个参与实际正在进行的工程开发工作的机会，而且最终的项目将使地球科学更容易为公众所利用。

The Earth System Curator collaboration will unify the treatment of models and datasets relating to climate change by developing a common language - a metadata formalism - with which to describe the two, and by prototyping a set of tools based on that formalism that allows researchers to manipulate models and datasets seamlessly and with ease. The goal, in the end, is to increase the productivity of climate researchers and understanding of the Earth system.Collaborators include computer science and Earth science researchers at MIT, Princeton University, the Georgia Institute of Technology, and the National Center for Atmospheric Research. The work proposed builds on two ongoing community efforts, the Earth System Modeling Framework (ESMF) and the Earth System Grid II (ESG). The ESMF is a national initiative to develop common modeling infrastructure for the nation's climate and weather models, including coupling tools and standard modeling utilities. The primary objective of ESG is to make the output of high-resolution, long-duration simulations performed with climate models available to global change impacts researchers nationwide, through the use of Grid, data/metadata, and portal technologies. The team will explore those aspects of ESMF and ESG that can be usefully aligned, and will prototype a new entity, the Earth System Curator, that spans the gap between the two.The Curator begins with a crucial insight: that the descriptors used for comprehensively specifying a model configuration are needed for a scientifically useful description of the model output data as well. The development of a common metadata schema that describes both will be the basis for this unique and powerful community resource. The Curator will provide a community database from which researchers can archive and query a wide class of Earth system models, experiments, model components, and model output data and results. Researchers will subsequently be able either to analyze model output from pre-existing runs, or to access a model and modify and run it themselves, either on a local computer or on the virtualized resources of the computational Grid. In addition to the query function, the project will prototype a tool that will test if sets of model components or datasets can interact to form an application. Finally, as part of the Curator effort, tools for auto-generation of component wrappers and applications will also be prototyped.The Curator is part of a community vision for the use of information technology in climate and related research. The Curator prototype will help to further suggest and define the form of next-generation modeling and data management tools, by offering a concrete representation to add credibility to innovative ideas. Further, the ESMF and ESG co-investigators, by virtue of their projects' emphases on production software and extensive customer bases, are in an excellent position to transition the Curator tools into a viable product following an NSF-funded prototype stage. Over the course of the Curator effort, many researchers associated with ESMF and ESG will be encouraged to try out and offer feedback on the Curator software. Advances achieved with the Curator project will influence other domains through conferences and publications, and through a web of relationships founded on a shared need for multi-component HPC modeling, ease of information archival and access, and similarities in simulation numerics. While the advances in climate prediction due directly to the Curator effort itself may be both difficult to track and modest, an integrated environment for Earth system research is critical to addressing world climate issues in the near future, and the Curator is a definitive step in that direction. The Curator will also be incorporated into the project infrastructure for software engineering courses at Georgia Tech. Appropriate introductory material will be prepared and project opportunities defined. Many of the project opportunities will take the form of making climate data accessible to the general public. In this way the Curator not only provides software engineering students an opportunity to participate in an actual ongoing engineering development effort, but the resultant projects will make Earth science more available to the general public.