Data Terra: a federated research infrastructure transforming Earth system data into knowledge and services for science and public decision-making

Data Terra is the French national research infrastructure dedicated to the observation, understanding and monitoring of the Earth system, with the explicit objective of transforming qualified Earth observation data into knowledge, services and indicators supporting scientific research, Earth system digital twins and public decision-making. It federates several long-standing thematic data and service hubs covering atmosphere, ocean, continental surfaces, solid Earth, biodiversity and provides high-resolution Earth observation. Together, these thematic poles curate, qualify and disseminate reference datasets, often based on long, homogeneous time series, essential to address major scientific challenges related to climate change, environmental dynamics and natural hazards.

A core objective of Data Terra is to foster interdisciplinarity through enhanced interoperability across disciplines, data types and scientific communities, a prerequisite for integrated Earth system science. This approach is strongly aligned with international frameworks of Essential Variables (Climate, Ocean, Land and Biodiversity EV), providing a shared scientific backbone for data production, qualification and reuse. The thematic poles structure their datasets, services and indicators around these Essential Variables, ensuring scientific consistency while enabling cross-domain analyses.

Data Terra relies on a federated, interoperable and scalable model, designed to be deployed and reused at different organisational and geographical scales. Its architecture and governance enable interoperability between national thematic poles, as well as integration with European and international initiatives, notably as a potential thematic or national node within the European Open Science Cloud (EOSC). This multi-scale design allows data, services and workflows developed within Data Terra to be exposed, combined and reused in broader research infrastructures without duplication or loss of semantic coherence.

Beyond core scientific use, Data Terra explicitly targets downstream applications such as Earth system digital twins, environmental services and decision-support tools for public policies. To strengthen the connection between scientific production, territorial needs and decision-makers, Data Terra has established regional and thematic coordination mechanisms (ART – Animations Régionales Thématiques). These ARTs act as interfaces between researchers, public authorities, private stakeholders and end-users, supporting the co-construction of indicators, dashboards and operational products adapted to policy and territorial contexts.

To support the full data-to-decision chain, Data Terra implements a coherent set of technical and semantic solutions. Semantic and machine-actionable interoperability is addressed through a pivot metadata model based on DCAT, combined with a shared repository of semantic artefacts, including controlled vocabularies, concept schemes and mappings. This enables automated discovery, cross-domain navigation and integration across platforms and infrastructures.

Technical interoperability relies on widely adopted standards and protocols, including OGC APIs for data discovery and access, S3-compatible object storage and cloud-optimised data formats such as ARCO. Emphasis is placed on the portability and reproducibility of data processing workflows, enabling execution across heterogeneous and federated computing environments.

Finally, Data Terra simplifies user interaction with complex and heterogeneous datasets to maximise scientific and societal impact. This is achieved through integrated resource catalogues linking datasets with example notebooks and documented use cases, advanced data preview and code generation capabilities, the federation of computing resources, and the development of dashboards and indicators.