Data and Convergence Science

Data are representations of phenomena and properties that assume a specific description and theory of the world. They are inherently context-dependent and are shaped by the people who collect them, their purpose and intended use, the practices used to collect them, prevailing norms, background beliefs, and assumptions made. When we are missing information about data’s context, we risk misunderstanding what data actually represent, choosing inadequate data for our purposes, or incorrectly relating different datasets. This is especially true for convergence science, which involves integrating multiple, different datasets – often from disparate areas of expertise – which can lead to a compounding of data-related issues. 

Using data for convergence research requires the same thoughtful planning, careful implementation, and rigorous adherence to established data standards and practices that all high quality research demands.  Research teams should have conversations about their data and analysis plans early and should continue discussing and revising plans throughout the entire research process. These conversations should always stem from and tie back to the critical, complex convergence science problem being addressed. Any decisions made around data and the reasons behind those decisions should be documented. All team members should have a basic understanding of the data being used by the team, even if they are not collecting or analyzing the data themselves.  

The NSF NCAR Convergence Science Program (CSP) has formed a working group devoted to issues related to data and convergence science.  This group has compiled important data-related considerations and best practices related to the topics listed below to support researchers working with convergence. As this work continues, the content on these pages will be refined and new resources will be added.

This resource was developed by NSF NCAR’s Convergence Science Program with contributions from (in alphabetical order by last name): Mariana Cains, Julie Demuth, Joseph Gum, Daniel Howard, Kimberly Fewless, Monica Morrison, Andrea Schumacher, Tanya Vance, Jeff Weber, Olya Wilhelmi.  Please contact csp_leadership@ucar.edu if you have any questions, feedback, or ideas.

existing-img-2.png

Using Existing Data for Convergence Research

Questions to consider when making decisions about what data to use for convergence research

integration-img.jpg

Data Integration

Important considerations and best practices for using data from disparate areas of expertise to address vexing problems in convergence research

sharing-img.jpg

Documenting and Sharing Data

Making sure we are documenting and sharing our data's story in a way that facilitates its use in convergence research

resources-img.jpg

Data-Related Examples and Resources

Examples of how NSF NCAR scientists are working with data for convergence research, planning templates, and more!

Want to learn more about Convergence Science?

What is convergence research?: A 2-page summary provided by the NSF NCAR CSP

NSF NCAR Convergence Science Program Clearinghouse: Collects, categorizes, and distributes convergence-related opportunities, which includes links to data-related repositories, resources, and publications.