An initiative of the European Commission

Community - Working Group

Earth Sciences

Purpose/Scope

The Earth Sciences Working Group investigates how AI can improve our understanding of Earth’s complex systems and support evidence-based responses to climate and environmental challenges.

The group covers climate science, Earth observation, agriculture, biodiversity, ecosystems, water resources and urban environments. It promotes collaboration between Earth scientists, AI researchers and infrastructure providers to develop trustworthy AI solutions that integrate diverse environmental datasets, improve predictive modelling and strengthen Europe’s leadership in climate resilience and sustainable resource management.

Key Questions

01

How can AI improve climate and environmental prediction?

Developing robust models for climate forecasting, environmental monitoring and risk assessment.

02

How can AI integrate heterogeneous Earth observation data?

Exploring methods for combining satellite, sensor and in situ observations.

03

How can AI support sustainable resource management?

Applying AI to agriculture, biodiversity conservation and ecosystem monitoring.

04

How can AI models remain transparent and trustworthy?

Ensuring interpretability and scientific confidence in environmental decision support.

Linked RAISE Forum Space

This discussion space connects researchers working on AI applications across the Earth sciences.

Members can exchange datasets, methods, infrastructure developments and research priorities while fostering interdisciplinary collaboration on climate and environmental challenges.

Experts

Selected experts contributing to the AI in Science Working Groups

Each Working Group brings together a selected group of experts contributing to the co-creation of the AI in Science Strategic Research and Innovation Agenda (SRIA). Members were appointed through an open call for expression of interest based on their expertise, experience, and ability to contribute to interdisciplinary collaboration across AI and scientific research domains.

Mariana Belgiu

Working Group Co-Chair

UT-ITC
Netherlands

Ioannis Athanasiadis

Wageningen University & Research
Netherlands

Paolo Ruti

EUMETSAT
Germany

Carlo Buontempo

ECMWF
Germany

Petteri Vihervaara 

Finnish Environment Institute (Syke)
Finland

Oliver Phillips

University of Leeds
United Kingdom

Rene Heim

University of Göttingen
Germany

Jian Peng

Helmholtz Centre for Environmental Research, UFZ
Germany

Stef Lhermitte

KU Leuven
Belgium

Sabine Griessbach

Jülich Supercomputing Centre, Forschungszentrum Jülich GmbH
Germany

Devaraju Narayanappa

CSC, IT center for science
Finland

Mikk Espenberg

University of Tartu (Estonia); University of Aberdeen (UK)
Estonia

Masahiro Ryo

Leibniz Centre for Agricultural Landscape Research (ZALF)
Germany

Fakhereh Alidoost

eScience Center
Netherlands

Vincent Miele

CNRS
France

Italo Epicoco

Euro Mediterranean Center on Climate Change
Italy

Events

  • Good Practices
  • AI Science Community
  • Earth Sciences
  • AI Research

CY-Bench

A benchmark dataset for sub-national crop yield forecasting

  • Good Practices
  • AI Science Community
  • Earth Sciences
  • AI Research

ESA WorldCover 2021

Machine-learning production and independent validation of a global land-cover map at 10 m resolution

  • Good Practices
  • AI Science Community
  • Earth Sciences
  • AI Research

MLCast

Open-source community infrastructure for machine-learning weather nowcasting

  • Documents
  • AI Science Community
  • Earth Sciences
  • Materials Science

AI in Materials Science​

SCIANCE is gathering evidence on scientific challenges, AI capabilities, infrastructures and emerging practices in “AI in Materials Science​”​, supporting the identification of strategic priorities for RAISE.

  • Documents
  • AI Science Community
  • Earth Sciences

AI in Earth Sciences​

SCIANCE is gathering evidence on scientific challenges, AI capabilities, infrastructures and emerging practices in “AI in Earth Sciences​”​, supporting the identification of strategic priorities for RAISE.

  • Good Practices
  • AI Science Community
  • Earth Sciences
  • AI Research

WorldCereal

Open, retrainable crop-mapping workflows with emerging foundation-model integration

  • Good Practices
  • AI Science Community
  • Earth Sciences
  • AI Research

YieldSAT

A multimodal benchmark dataset for field and subfield crop yield prediction

  • Good Practices
  • AI Science Community
  • Earth Sciences
  • AI Research

Tessera

Open reuse of geospatial foundation-model embeddings for Earth observation

  • Good Practices
  • AI Science Community
  • Earth Sciences
  • AI Research

BigEarthNet

Large-scale data foundation for AI in Earth observation

Other Working Groups