The EGU General Assembly 2027 will bring together geoscientists from around the world for a week of scientific exchange, collaboration, and discovery across all fields of the earth, planetary, and space sciences.
From 4–9 April 2027, the international geoscience community will reconvene at the Austria Center Vienna (ACV) and online, continuing EGU’s tradition of providing an open, inclusive, and vibrant forum for sharing the latest research, discussing emerging challenges, and fostering new collaborations.
Explore Relevant Sessions
The ENVRI community is excited to support and promote several key sessions that highlight the collaborative role of RIs in advancing environmental science.
Deadline for submissions: Tuesday, 12 January 2027, 13:00 CET
Session ESSI2.8 – Machine and AI actionable FAIR Digital Objects enabled VRE’s, for Environmental and Earth Science
Convener: Jacco Konijn | Co-conveners: Anne Fouilloux, Marie Jossé, Alessandro Rizzo, Eugenio Trumpy
Environmental science, scientists and developers increasingly use or build Artificial Intelligence in tools, Digital Twins and models, but are concerned about trust, transparency and digital sovereignty when supporting Open Science practices. AI-ready research assets and VRE’s are crucial enablers.
Fair Digital Objects (FDO)’s become the technical backbone that makes AI trustworthy enough to use. Agentic search is the clearest case for this: an agent querying across catalogues can only return trustworthy results if the underlying objects carry typed metadata, provenance and clear licensing.
Research Infrastructures, providing the standards for FDO’s and trustworthy data are at the center of these developments and have a responsibility to do a risk and value assessment before any AI tool is adopted, model cards, transparency, and shared European compute rather than dependence on commercial cloud providers.
Virtual Research Environments (VRE) and workflow builders provided by Research Infrastructures need to take these requirements into account and provide solutions that are machine actionable while at the same time are trustworthy for any AI application.
We envision Virtual Research Environments (VREs) that are not only machine-actionable but also deeply trustworthy, designed to integrate AI assistance while upholding FAIR principles, digital sovereignty, and Open Science standards. These VREs should empower scientists to safely explore AI-driven opportunities -whether in data sharing, integration, or analysis- with full confidence in the tools, data, and workflows provided. By embedding provenance, interoperability, and ethical AI practices, we can ensure that VREs become catalysts for collaborative, reproducible, and impactful environmental research.
This session aims to bring together practitioners, developers and stakeholders to discuss the technical challenges and share practices with the aim to contribute to the emerging ENVRI & Data Terra/EOSC nodes.
Session ITS5.6/BG8.9 – From Essential to Actionable: Linking Research Infrastructures, Earth Observations and Climate Services
Convener: Sindu Raj Parampil | Co-conveners: Theresia Bilola, Anca Hienola
Essential Variables (EVs) provide a common framework for characterising key components of the Earth system, but their potential extends beyond observation and monitoring. This session explores how EV frameworks can connect heterogeneous observing systems, from Research Infrastructures (RIs) such as ICOS and ACTRIS, to satellite data and models, and support the transformation of essential observations into actionable climate information.
European Environmental Research Infrastructures (ENVRI) are a key part of this integration layer. Beyond providing observations, RIs contribute long-term stewardship, calibration and quality assurance, standards, provenance, data and access services, computational environments and increasingly integrated information services. Together with satellites and models, these capabilities provide operational foundations for connecting observations across domains and scales.
Session SC2.18 – Build your own workflow in the Cloud!
Convener: Jacco Konijn | Co-conveners: Zhiming Zhao, Eleonora Parisi, Koen Greuell, Cosimo Vallo
Earth and environmental sciences thrive on data diversity: from ocean temperatures to biodiversity records, from climate indicators to ecological observations and their interactions in ecosystems or relation to geophysical circumstances. Yet, this very diversity can also be a barrier: different datasets are described with different standards and stored in different formats.
Research Infrastructures (RI’s) provide seamless access to environmental datasets, like the Environmental Research Infrastructure Community (ENVRI) that has developed the ENVRI-Hub. On top of that, RI’s offer Vvirtual Rresearch Eenvironments as a service where scientists can integrateturn their own code into workflows and execute these workflows with this data in the cloud .
This short course will give researchers a practical introduction to how workflow builders like NaaVRE, as developed by LifeWatch ERIC, can directly support their own projects to reduce knowledge loss and build more reproducible and impactful science.
Check out the training resource from our EGU26 Short Course

