Agricultural research data are valuable resources that are often lost after the original research objectives are complete. Numerous datasets have become findable and accessible through data portals such as CGIAR’s GARDIAN and the USDA’s Ag Data Commons but many of these legacy research datasets are difficult for researchers to reuse due to the diversity of content, vocabularies, units, domains, and schemas. ARDN provides dataset annotation protocols that allow quantitative agronomic data to be machine-interpreted and translated to a common format with a standardized vocabulary. The original raw dataset, often associated with a DOI, are maintained without modification but are linked to the new interoperable data products. Using these data annotation protocols, a core set of interoperable agronomic data from multiple datasets can be discovered, harmonized, and aggregated into new data products for quantitative research including modeling, data analytics, and meta-analyses.

These ARDN protocols expand upon data interoperability standards developed as part of the Agricultural Model Intercomparison and Improvement Project (AgMIP) and are being implemented in Ag Data Commons; GARDIAN Labs as part of the CGIAR’s Platform for Big Data; and at the International Fertilizer Development Center (IFDC) as part of their open data efforts.

Presenter(s)

Cheryl Porter