Researchers at the Swiss Federal Institute of Technology Zurich have successfully transferred approximately 100 petabytes of publicly available NASA climate and environmental data to the Swiss National Supercomputing Centre in Lugano. The operation, which required roughly one year to complete, involved migrating nearly six billion files to safeguard a critical repository of Earth observation records. ETH Zurich announced the initiative late last week, with researchers confirming plans to subsequently ingest comparable datasets from the U.S. National Oceanic and Atmospheric Administration. The transfer was initiated by ETH professor and former NASA chief scientist Thomas Zurbuchen and spearheaded by Reto Knutti, head of the university’s Center for Climate Systems Modeling. While the dataset remains legally accessible, the move was partially motivated by recent federal budget reductions to U.S. climate science and Earth observation programs under the current administration. Knutti emphasized that although no access restrictions have been implemented to date, the rapid pace of U.S. policy shifts necessitates proactive data preservation. He characterized the archived information as foundational intellectual property, noting that sustained global monitoring depends entirely on decades of continuous measurement campaigns. Once secured, the data will power advanced artificial intelligence training pipelines at the supercomputing facility. Traditional physics-based climate models rely on complex mathematical simulations and typically require hours to generate global forecasts. In contrast, AI foundation models are already demonstrating superior predictive accuracy and can process planetary-scale weather projections in under a minute, achieving speeds up to a thousand times faster than conventional methods. Rapid, reliable forecasting is critical for disaster preparedness, agricultural planning, and hydropower management. By integrating the NASA archive with the Alps supercomputer, one of the world’s most powerful computing systems, researchers aim to accelerate pattern recognition and improve long-term climate impact assessments. The collaboration underscores a growing reliance on data-driven scientific infrastructure to mitigate climate risks. ETH officials stressed that the primary objective is analytical utility, ensuring that massive volumes of environmental metrics, including greenhouse gas concentrations, cloud formations, precipitation patterns, and ice sheet dynamics, are actively processed rather than passively stored. With machine learning capabilities advancing rapidly, the combined storage and computing resources will enable more granular hazard modeling and adaptive climate response strategies. The initiative reflects a broader trend in scientific research where geographic and political considerations increasingly intersect with digital preservation and computational forecasting.