Making STAC FDO-ready: A Practical Path toward FAIR Digital Objects in Geoscientific Data Spaces

FAIR Digital Objects (FDOs) provide an actionable framework for implementing the FAIR principles by combining persistent identifiers with machine-readable metadata, explicit typing, and structured relations. The FDO Forum, as an open, community-driven initiative, develops and coordinates specifications and reference concepts to support interoperable digital objects across infrastructures. A key challenge, however, is demonstrating how these specifications can be applied in practice within existing data ecosystems, where established domain standards and evolving collections must be integrated rather than replaced.

In this contribution, a practical implementation of FDO specifications is presented using the SpatioTemporal Asset Catalog (STAC) as an example. As a widely adopted standard for spatio-temporal data, STAC's modular design makes it an ideal bridge between established community practices and the FDO paradigm. The demonstration shows how STAC objects are transformed into typed FDOs using Handle-based PIDs and registered object types via a Data Type Registry (DTR). This approach enables machine-actiolnable navigation and interpretation that transcends domain-specific tooling.

The approach is illustrated using a STAC-based catalog developed at the German Climate Computing Center (DKRZ), reflecting typical characteristics of climate research and climate modelling data, such as evolving and versioned collections and multiple levels of aggregation. The focus is on the practical application of FDO specifications, illustrating how typing, identifiers, and relations can be introduced in a standards-compliant manner without disrupting existing infrastructures, while enabling stable referencing, automated discovery, and seamless integration into data-processing workflows.

The results show that implementing FDO specifications through STAC is a pragmatic and transferable pathway from specification-level concepts to operational adoption. The implementation enables the creation of interoperable, machine-actionable data spaces while building on established standards and tooling, and provides lessons learned for other infrastructures aiming to operationalize FAIR Digital Objects in practice.