Hackathon hosted by the OpenGeoHub Foundation as part of the 2025 OpenGeoHub Summer School, designed to explore the potential of cutting-edge remote sensing embeddings for forest monitoring. This challenge invites students, early-career researchers, and GIS/EO practitioners to develop machine learning models that predict forest canopy height from Google DeepMind’s AlphaEarth Foundations (AEF) embeddings, derived from Sentinel-1, Sentinel-2, and Landsat imagery.

Participants will compete to benchmark the accuracy and generalizability of AEF embeddings against GEDI-derived canopy height measurements, gaining hands-on experience with real-world data, open-source tools, and environmental AI workflows.

The winning team will present their results at the 2025 OpenGeoHub Summer School—a non-profit event bringing together developers, scientists, and institutions committed to open environmental data and software. The 2025 edition is co-organized with EO Council, Netherlands Space Office, and SURF, and includes live demos, workshops, panel discussions, and social gatherings.

Hackathons are supported by the EU-funded Open-Earth-Monitor project, under the Horizon Europe research and innovation programme (grant agreement No. 101059548).