Solaris: A Foundation Model of the Sun
Not on the map. Why?
Abstract (arXiv)
Foundation models have demonstrated remarkable success across various scientific domains, motivating our exploration of their potential in solar physics. In this paper, we present Solaris, the first foundation model for forecasting the Sun's atmosphere. We leverage 13 years of full-disk, multi-wavelength solar imagery from the Solar Dynamics Observatory, spanning a complete solar cycle, to pre-train Solaris for 12-hour interval forecasting. Solaris is built on a large-scale 3D Swin Transformer architecture with 109 million parameters. We demonstrate Solaris' ability to generalize by fine-tuning on a low-data regime using a single wavelength (1700 AA), that was not included in pre-training, outperforming models trained from scratch on this specific wavelength. Our results indicate that Solaris can effectively capture the complex dynamics of the solar atmosphere and transform solar forecasting.
Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Builds on 10 of its 15 references that are in the map. First posted on arXiv 25 Nov 2024.
Builds on
10 of 15 references.
References · 10
- 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
- 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,139
- 2021On the Opportunities and Risks of Foundation Modelscited by 4,033
- 2021Swin-Unet: Unet-like Pure Transformer for Medical Image Segmentationcited by 1,222
- 2021Perceiver: General Perception with Iterative Attentioncited by 1,104
- 2023ClimaX: A foundation model for weather and climatecited by 337
- 2022GraphCast: Learning skillful medium-range global weather forecastingcited by 267
- 2024Poseidon: Efficient Foundation Models for PDEscited by 200
- 2022ChemBERTa-2: Towards Chemical Foundation Modelscited by 127
- 2019A Machine Learning Dataset Prepared From the NASA Solar Dynamics Observatory Missioncited by 26
Built on by
Citing papers per year: 2025: 4.
4 papers.
- 2025Surya: Foundation Model for Heliophysicscited by 6
- 2025Predicting partially observable dynamical systems via diffusion models with a multiscale inference schemecited by 5
- 2025Graph-based Neural Space Weather Forecastingcited by 1
- 2025Contrastive Heliophysical Image Pretraining for Solar Dynamics Observatory Recordscited by 0
- 2025Contrastive Heliophysical Image Pretraining for Solar Dynamics Observatory Records28 Nov 2025
- 2025Predicting partially observable dynamical systems via diffusion models with a multiscale inference scheme24 Nov 2025
- 2025Graph-based Neural Space Weather Forecasting23 Sep 2025
- 2025Surya: Foundation Model for Heliophysics18 Aug 2025