发表机构
Argonne National Laboratory; University of Illinois Chicago; Northwestern University（阿贡国家实验室; 伊利诺伊大学芝加哥分校; 西北大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
OpenCosmo项目旨在为大型宇宙学模拟数据集提供灵活访问和分析模式，通过网络数据门户和分析库，结合现有设施支持多层次交互，基于Globus Compute的架构可扩展，助力宇宙学数据共享与计算结合。
AI 中文摘要
宇宙学是一门观测精密科学，详细且富含科学内容的模拟是诸多分析的必要组成部分。这些模拟运算成本高且产生大量复杂数据。广泛共享数据以促进进一步探索、与观测对比及与大众交流对推动科学前沿和吸引更广泛群体至关重要。本文介绍OpenCosmo项目，旨在为大型宇宙学模拟数据集提供灵活访问和分析模式。提供易用的基于网络的数据门户来检索旗舰级宇宙学数据集的可下载子集，还有复杂分析库用于对返回数据进一步分析。通过与现有高性能计算和数据基础设施集成，支持从简单搜索下载到交互式探索计算等多层次交互。其基于Globus Compute的架构提供了可扩展且适应性强的框架，可扩展到其他寻求将数据共享与计算能力结合的科学领域。
英文摘要
Cosmology is a precision observational science, and large simulations are necessary components of many analyses. These simulations are computationally expensive and produce massive, complex datasets; sharing them widely -- to enable further explorations, comparison with observations, and communication with general audiences -- is crucial to realizing their scientific value. In this paper, we introduce the OpenCosmo project, which provides flexible access to, and analysis of, flagship cosmological simulations performed with HACC. A web-based portal (https://opencosmo.science) serves custom subsets -- halo catalogs, profiles, particles, galaxy catalogs, and lightcone catalogs and maps -- from simulations including the two-trillion-particle Frontier-E gravity-only run, Last Journey, Discovery, and a 64-member hydrodynamic suite. A companion Python toolkit analyzes the returned data and scales without modification from laptop-sized subsets to full simulations on supercomputers. OpenCosmo supports multiple levels of interaction, from browser-based search and download to programmatic and AI-agent-driven workflows, by integrating with existing high-performance computing and data infrastructure. Its architecture, built on Globus services, provides a scalable and adaptable framework that can be extended to other scientific domains seeking to couple data sharing with computational capability.
Comments18 pages, 6 figures. Submitted to the Open Journal of Astrophysics
