| The Arctic is changing rapidly with dramatic local and global effect. To understand that change requires understanding the Arctic as a system. Models of different processes and at various scales are necessary tools for analyzing and understanding the Arctic system. Models are extremely diverse, yet they all require quality data. Through a series of case studies, augmented with with ethnographic observation around the International Polar Year, this thesis examines how modelers assess, acquire, and prepare data for their models. By comparing specific case studies, common themes emerge that can be compared against broader observation. These themes, in turn, suggest data management techniques or requirements for data systems to improve access and use by modelers and generally improve understanding of the Arctic system. This case study based approach has proven to be a useful method for teasing out both general and specific data needs for different models. An overarching lesson is that greater short-term benefit to modelers and significant gains in efficiency can be achieved by improving the formats, convention, and consistency of the data rather than improved interfaces and analysis tools. A "data-first" philosophy can improve the data systems that support the overall interdisciplinary, integrative science necessary to understand the complex Arctic system. |
