Data Policies
NSF NCAR’s Geoscience Data Exchange (GDEX) is committed to providing high-quality, reliable Earth system data with comprehensive documentation to support informed user decisions about appropriate applications. We ensure data are:
- Quality-assured through rigorous validation and verification processes
- Well-documented with detailed metadata and provenance to enable users to assess the fitness of data for their research purposes
- Discoverable through standardized protocols and metadata
- Accessible through resilient, maintained infrastructure
- Transparently described regarding data limitations, processing methods, and design contexts
We recognize that data fitness is determined by individual user needs and applications. Our commitment is to provide the documentation, quality metrics, and contextual information necessary for users to make informed judgments about data suitable for their intended uses.
GDEX is NSF NCAR’s modern community data commons, purpose-built to support AI/ML workflows, large-scale analysis, and reproducible science. GDEX provides a resilient, policy-driven foundation that ensures data are not only accessible, but resilient, reliable, well-described, and fit for advanced computational use.
Designed for AI & Advanced Analysis
GDEX prioritizes analysis-ready and AI-ready datasets, emphasizing standardized formats, rich metadata, and colocated access with NCAR’s community compute resources. This reduces friction for users running machine-learning, statistical-analysis, and high-performance workflows, while supporting rapid iteration and reproducibility.
CoreTrustSeal Certification
CoreTrustSeal is an international, community-driven certification that defines what it means for a data repository to be resilient, reliable, and sustainable. Research data repositories widely use the certification to demonstrate responsible long-term stewardship of digital assets.
At a high level, CoreTrustSeal is organized around three pillars: Organizational Infrastructure, Digital Object Management, and Technology & Security.
This ensures that datasets are curated, maintained, and governed in accordance with internationally recognized best practices, critical for scientific integrity and downstream AI use.
FAIR & Open by Design
FAIR data principles are internationally adopted guidelines that ensure research data are Findable, Accessible, Interoperable, and Reusable. FAIR does not mean “open by default,” but rather that data are well-described, discoverable, and usable, by both humans and machines.
FAIR provides the foundation that allows platforms like Geoscience Data Exchange (GDEX) to support scalable, trustworthy, and future-ready data use by enabling:
- Machine-actionable discovery and access
- Reliable data pipelines for AI/ML workflows
- Reproducible and auditable scientific results
- Long-term value beyond a single research project
All GDEX datasets follow FAIR data principles through persistent identifiers, standardized metadata, clear licensing, and open access policies. FAIR alignment enables datasets to be easily discovered, combined across domains, and reused in new scientific and AI contexts.
link