Speaker
Description
Modern digital infrastructures are critical for transforming long-term ocean observing systems into scalable, accessible assets for marine research and the blue economy. This contribution presents the architectural framework and technological evolution of the digital data infrastructures and operational services deployed by the Spanish Institute of Oceanography (IEO-CSIC), acting as a National Oceanographic Data Center (https://nodc.ieo.es/geonetwork), as well as the developments during the project Integrated Platform of the Marine Sciences Complementary Plan in Galicia (https://ccmmbigdata.cesga.es).
Operating sustained observing systems since the 1980s, IEO has transitioned toward robust, open-source-based architectures to ensure that massive, heterogeneous environmental datasets comply strictly with FAIR principles. The underlying digital pipeline ingests data formats (such as CNV and NetCDF) using standardized data models. Metadata management is anchored in ISO 19139.sdn.csr and Cruise Summary Report (CSR) standards, enriched with controlled vocabularies from SeaDataNet and the British Oceanographic Data Centre (BODC).
The backend relies on PostgreSQL + PostGIS for spatial data storage and MinIO for scalable object storage. To support high-performance cloud-native workflows, the infrastructure incorporates modern formats like Parquet and Zarr, and there is a development of STAC (SpatioTemporal Asset Catalogs) powered by pgSTAC and stac-fastapi-pgstac for rapid geospatial discovery. Dissemination is driven by hybrid services: GeoNetwork powers the geographic catalog node, while GeoServer (acting as the IEO Spatial Data Infrastructure node) delivers classic OGC web services (WMS, WFS, CSW) alongside emerging OGC-API standards.
Furthermore, during the development of the integrated marine data platform in Galicia, the infrastructure was expanded beyond traditional oceanography data to handle complex multimedia workflows—specifically managing and processing phytoplankton images captured via FlowCam and submarine videos. Utilizing Python and Jupyter Notebooks, the system implements taxonomic standardization through the World Register of Marine Species (WoRMS), links multimedia content with tabular records, and extracts structural features using computer vision edge detection algorithms (such as Canny).
In summary, this contribution highlights how combining advanced cloud-native geospatial architectures, STAC catalogs, and multimedia-enabled digital services provides a scalable paradigm for modern marine research infrastructures in the North-Atlantic region.