Advancing Weather Forecasting with Data Integration
New AI Model ESFM Predicts Global Extreme Weather
ETH Zurich researchers launch a foundation model integrating land, sea, and air data to forecast super typhoons.
A digital rendering of the Earth with glowing data lines representing the atmosphere, land, and water interacting to form a storm system.
Photo: Avantgarde News
Scientists from ETH Zurich and the Swiss AI Initiative have launched the Earth System Foundation Model (ESFM) [1][2]. This new artificial intelligence framework autonomously learns interactions between the atmosphere, land, and water [1]. Unlike previous models that only analyzed the air, ESFM combines diverse datasets to improve precision [1][3].
The system predicts extreme events like super typhoons with high accuracy [1]. It is designed to work even when satellite or ground station data is missing [1][2]. This capability allows the model to fill critical information gaps in global weather forecasting [2].
Researchers state this unified framework represents a major step in data integration [3]. By understanding how different earth systems work together, the model shows exactly how extreme weather originates [2]. This helps scientists prepare for climate-related disasters more effectively.
Editorial notes
Transparency note
AI assisted drafting. Human edited and reviewed.
- AI assisted
- Yes
- Human review
- Yes
- Last updated
Risk assessment
Reviewed for sourcing quality and editorial consistency.
Sources
- 1.↗
eurekalert.org
https://www.eurekalert.org/news-releases/1127851
- 2.↗
ethz.ch
https://ethz.ch/de/news-und-veranstaltungen/eth-news/news/2026/05/neue-ki-schliesst-datenluecken-und-zeigt-wie-extremwetter-auf-der-erde-entsteht.html
- 3.↗
researchgate.net
https://www.researchgate.net/publication/404427307_Earth_System_Foundation_Model_ESFM_A_unified_framework_for_heterogeneous_data_integration_and_forecasting
Related stories
View allTopics
About the author
Avantgarde News Desk covers advancing weather forecasting with data integration and editorial analysis for Avantgarde News.
