ETH Zurich copies 100 petabytes of NASA climate data to Swiss supercomputer
This digest was compiled by AI from multiple sources — links to the originals are below.
Researchers at the Swiss Federal Institute of Technology Zurich (ETH) have copied around 100 petabytes of NASA climate and environmental data to the Swiss National Supercomputing Centre (CSCS) in Lugano. The transfer, which took about a year and involved roughly 6 billion files, aims to train AI models for weather and climate forecasting. The move also serves as a safeguard against potential U.S. funding cuts that have reduced federal climate science programs under President Donald Trump.
Data Transfer Scope
ETH Zurich announced the data transfer late last week, with researchers copying approximately 100 petabytes of NASA data to the CSCS in Lugano. Reto Knutti, professor of climate physics and head of ETH's Center for Climate Systems Modeling (C2SM), told AFP the volume is equivalent to about 20 million feature-length films. The datasets, gathered over decades, include information on greenhouse gases, clouds, precipitation, and ice sheets.
Motivation and Concerns
While the primary goal is to train AI models for rapid weather and climate forecasting, the researchers acknowledged concern over U.S. funding cuts. ETH professor and former NASA chief scientist Thomas Zurbuchen emphasized the United States' historical role in Earth observation. Knutti noted that so far, U.S. access to data has not been restricted, but decisions in the administration can be quick and unpredictable.
AI Model Potential
Knutti described data-driven foundation models as already outperforming traditional physics-based models in prediction skill, while being up to 1,000 times faster. Such models can produce a global weather forecast for multiple days in about a minute, compared to hours with conventional methods. The researchers also plan to copy large datasets from the U.S. National Oceanic and Atmospheric Administration (NOAA) in the future.
What's Next
The researchers plan to continue copying datasets from NOAA to expand the AI training repository. It remains unclear whether future U.S. policy changes could restrict access to additional data, potentially affecting global climate monitoring efforts.
1 source
ETH Zurich copies 100 petabytes of NASA climate data to Swiss supercomputer
