Guided Unconditional and Conditional Generative Models for Super-Resolution and Inference of Quasi-Geostrophic Turbulence

2025

Read Guided Unconditional and Conditional Generative Models for Super-Resolution and Inference of Quasi-Geostrophic Turbulence (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-2507-00719.txt.

Description

Anantha Narayanan Suresh Babu and co-authors apply diffusion models to sharpen coarse views of ocean-like turbulence. Earth system simulations are coarse, and observations are sparse with gaps. Quasi-geostrophic turbulence is a standard test bed for both flaws.

Four generative approaches are compared on the same tasks. Two guide a pretrained model without new paired training. Two others train on paired high and low detail data. Tests span eddy flow and jet flow regimes. They also span two flow strengths and two observation types. Scores cover field accuracy and turbulence statistics. Ensemble spread is checked for honest uncertainty. The study maps which method suits which data problem.

Catalog record

YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE88 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-00719

Browse the full library · About the Tower of Babel