A Survey of Foundation Models for Environmental Science

Runlong Yu 1, Shengyu Chen 1, Yiqun Xie 2, Xiaowei Jia 1 · 2025

Read A Survey of Foundation Models for Environmental Science by Runlong Yu 1, Shengyu Chen 1, Yiqun Xie 2, Xiaowei Jia 1 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-03142.txt.

Description

Runlong Yu and co-authors survey in 2025 how foundation models are being used across environmental science. Ecosystems are complex, tightly linked, and often short on data, so classic models strain to capture them. Foundation models, pretrained at large scale, can merge many data types, track change over space and time, and adapt to many tasks.

The survey covers uses in forward prediction, data generation, filling gaps from observations, sharpening coarse maps, blending model outputs, and decision making. It also walks through how such models are built: gathering data, picking a design, training, tuning, and testing. Work from weather, water, land, and ecology all appears. The authors aim to spark teamwork between machine learning builders and environmental scientists, arguing that shared tools could speed work on resource care and sustainable development.

Catalog record

CREATORRunlong Yu 1, Shengyu Chen 1, Yiqun Xie 2, Xiaowei Jia 1
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE46 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-03142

Browse the full library · About the Tower of Babel