Research supported by CONCERTO was presented at the GREEN-AI workshop, held in Naples alongside the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD), which took place between 7-11 September 2026. CONCERTO was represented by its project partner, the Barcelona Supercomputing Center (BSC), whose researchers presented work exploring how machine learning applications in Earth sciences can become more sustainable through accessible and practical recommendations.

The paper highlights the importance of making guidance on sustainable machine learning easy to understand and apply. As machine learning becomes increasingly integrated into Earth sciences research, addressing its environmental footprint is essential to ensuring that these technologies contribute to scientific progress while making more responsible use of computational resources.

The GREEN-AI workshop provided a platform for discussing the sustainability of machine learning and its application across different scientific fields, including Earth sciences. It brought together researchers to exchange ideas on greener machine learning practices and the broader challenges of developing responsible AI. The wider ECML PKDD programme also featured discussions on fairness in machine learning and reinforcement learning, as well as AI ethics and governance.

For CONCERTO, this research aligns with the project's broader efforts to improve the representation of the terrestrial carbon cycle by combining Earth observation, process-based modelling and innovative data-driven approaches. Considering the sustainability of machine learning methods is an important part of advancing these technologies responsibly and supporting their effective application in Earth system science.

The presented work represents an initial step towards more sustainable machine learning practices in Earth sciences and aims to encourage further discussion and collaboration within the research community. It was funded by CONCERTO, TerraDT and ELLIOT and carried out at the Earth Sciences Department of the Barcelona Supercomputing Center.