Earth System Foundation Model (ESFM): A unified framework for heterogeneous data integration and forecasting

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ozdemir, Firat, Cheng, Yun, Mohebi, Salman, Lehmann, Fanny, Adamov, Simon, Zhang, Zhenyi, Trentini, Leonardo, Grund, Dana, Fuhrer, Oliver, Hoefler, Torsten, Mishra, Siddhartha, Schemm, Sebastian, Soja, Benedikt, Salzmann, Mathieu
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915973207949312
author Ozdemir, Firat
Cheng, Yun
Mohebi, Salman
Lehmann, Fanny
Adamov, Simon
Zhang, Zhenyi
Trentini, Leonardo
Grund, Dana
Fuhrer, Oliver
Hoefler, Torsten
Mishra, Siddhartha
Schemm, Sebastian
Soja, Benedikt
Salzmann, Mathieu
author_facet Ozdemir, Firat
Cheng, Yun
Mohebi, Salman
Lehmann, Fanny
Adamov, Simon
Zhang, Zhenyi
Trentini, Leonardo
Grund, Dana
Fuhrer, Oliver
Hoefler, Torsten
Mishra, Siddhartha
Schemm, Sebastian
Soja, Benedikt
Salzmann, Mathieu
contents Foundation models (FMs) for the Earth system learn statistical relationships between physical variables across massive datasets to enable versatile downstream applications through finetuning, separating them from task-specific weather models. Here, we introduce Earth System Foundation Model (ESFM), a fully open model building on the 3D Swin UNet backbone of the pioneering Aurora model. ESFM introduces extensions that increase functionality and foster adoption in climate sciences. First, the encoding scheme and training protocols have been extended to handle diverse datasets, including those containing missing values across all spatio-temporal dimensions such as satellite data, as well as station data, all under one backbone. Axial attention is introduced to capture inter-variable dependencies. As a result ESFM skillfully predicts variables in regions or on pressure levels where no data is present at the initial time, while preserving inter-variable relationships, for example between temperature, pressure, and humidity. Individual variable tokenization enables different sets of variables to be shuffled during training and simplifies the process of building extensions for new downstream tasks. Adaptive layer norm-based ensembles allow for a simple yet effective way to transform deterministic ESFM to a probabilistic FM. We present findings using dense gridded data (ERA5, CMIP6), regionally masked dense data, sparse gridded MODIS satellite data, and station data. Results demonstrate competitive or superior performance relative to state-of-the-art benchmarks. Case studies of Super Typhoon Doksuri (2023) and 2024 sudden stratospheric warming events show accurate positional and magnitude estimations of extreme weather. ESFM retains the strengths of previous foundation models, such as long-term stability, but facilitates application to a variety of downstream tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2605_00850
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Earth System Foundation Model (ESFM): A unified framework for heterogeneous data integration and forecasting
Ozdemir, Firat
Cheng, Yun
Mohebi, Salman
Lehmann, Fanny
Adamov, Simon
Zhang, Zhenyi
Trentini, Leonardo
Grund, Dana
Fuhrer, Oliver
Hoefler, Torsten
Mishra, Siddhartha
Schemm, Sebastian
Soja, Benedikt
Salzmann, Mathieu
Atmospheric and Oceanic Physics
Artificial Intelligence
Machine Learning
Image and Video Processing
Foundation models (FMs) for the Earth system learn statistical relationships between physical variables across massive datasets to enable versatile downstream applications through finetuning, separating them from task-specific weather models. Here, we introduce Earth System Foundation Model (ESFM), a fully open model building on the 3D Swin UNet backbone of the pioneering Aurora model. ESFM introduces extensions that increase functionality and foster adoption in climate sciences. First, the encoding scheme and training protocols have been extended to handle diverse datasets, including those containing missing values across all spatio-temporal dimensions such as satellite data, as well as station data, all under one backbone. Axial attention is introduced to capture inter-variable dependencies. As a result ESFM skillfully predicts variables in regions or on pressure levels where no data is present at the initial time, while preserving inter-variable relationships, for example between temperature, pressure, and humidity. Individual variable tokenization enables different sets of variables to be shuffled during training and simplifies the process of building extensions for new downstream tasks. Adaptive layer norm-based ensembles allow for a simple yet effective way to transform deterministic ESFM to a probabilistic FM. We present findings using dense gridded data (ERA5, CMIP6), regionally masked dense data, sparse gridded MODIS satellite data, and station data. Results demonstrate competitive or superior performance relative to state-of-the-art benchmarks. Case studies of Super Typhoon Doksuri (2023) and 2024 sudden stratospheric warming events show accurate positional and magnitude estimations of extreme weather. ESFM retains the strengths of previous foundation models, such as long-term stability, but facilitates application to a variety of downstream tasks.
title Earth System Foundation Model (ESFM): A unified framework for heterogeneous data integration and forecasting
topic Atmospheric and Oceanic Physics
Artificial Intelligence
Machine Learning
Image and Video Processing
url https://arxiv.org/abs/2605.00850