Metraipolis › archive › dc_ad630dc36004
Iterative Refinement Diffusion for Super-Resolved Data Assimilation of Multiscale Physical Systems — arXiv:2608.14744
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.14744
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T19:39:52.002Z
- SHA-256 of the archived text
905a0f72e0e16e1c6ee82b156e62cc94f046d6923804a26eaabd09360dcf1570- Stable URL of this page
https://metraipolis.com/documento/dc_ad630dc36004
How to cite
Iterative Refinement Diffusion for Super-Resolved Data Assimilation of Multiscale Physical Systems — arXiv:2608.14744 — arXiv (en), https://arxiv.org/abs/2608.14744 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ad630dc36004 (Metraipolis archive dc_ad630dc36004, sha256 905a0f72e0e16e1c6ee82b156e62cc94f046d6923804a26eaabd09360dcf1570)
BibTeX
@misc{dcad630dc36004,
title = {Iterative Refinement Diffusion for Super-Resolved Data Assimilation of Multiscale Physical Systems — arXiv:2608.14744},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.14744},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ad630dc36004; sha256 905a0f72e0e16e1c6ee82b156e62cc94f046d6923804a26eaabd09360dcf1570},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.14744. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Recovering high-resolution states from sparse, low-resolution observations is a central challenge in scientific machine learning and data assimilation. Classical data assimilation exploits temporal information through forecast-analysis cycles, but often requires repeated access to expensive high-resolution forecast models. Generative super-resolution can recover unresolved structure from coarse observations, but is commonly used as a one-shot mapping that does not fully exploit constraints from past states. We introduce Iterative Refinement (IR), a learned data assimilation framework that combines these perspectives. Instead of performing a single coarse-to-fine reconstruction, IR decomposes the task into resolution-wise forecast-analysis operations across a multiresolution hierarchy. At each stage, a shared neural operator with resolution-dependent spectral mode slicing provides a dynamical prior, while a shared conditional diffusion corrector uses the current coarser-resolution state to produce a refined posterior at the next finer resolution. We evaluate IR on one-dimensional stochastically forced Burgers dynamics and two-dimensional Kraichnan turbulence. On the challenging 256x256 Kraichnan benchmark, IR achieves an RMSE of 0.184 and an SSIM of 0.836, outperforming spectral upsampling, one-shot diffusion super-resolution, enhanced deep super-resolution, and an autoregressive forecaster. On the more constrained Burgers testbed, IR remains competitive with one-shot diffusion, which achieves the lowest RMSE. These results show that one-shot generative reconstruction can be effective for simpler settings, while hierarchical forecast-analysis refinement becomes advantageous in strongly multiscale and underdetermined regimes. Overall, IR combines temporal priors, generative correction, and multiresolution reconstruction for learned data assimilation in complex physical systems. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-18 in cs.LG). Categorie arXiv: Machine Learning (cs.LG) ; Fluid Dynamics (physics.flu-dyn). Pagina del lavoro: https://arxiv.org/abs/2608.14744 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-18T19:02:44.048Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-08-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).