On the effectiveness of neural operators at zero-shot weather downscaling
Environmental Data Science · 2025 · 4
DOI: 10.1017/eds.2025.11 ↗Kurz gefasst
Fachartikel in der Zeitschrift Environmental Data Science, online erschienen am 27.03.2025. Verfasst von Saumya Sinha und 2 weiteren Autor:innen. Die Arbeit wurde bisher 11-mal zitiert (Crossref) und nennt 53 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Machine-learning (ML) methods have shown great potential for weather downscaling. These data-driven approaches provide a more efficient alternative for producing high-resolution weather datasets and forecasts compared to physics-based numerical simulations. Neural operators, which learn solution operators for a family of partial differential equations, have shown great success in scientific ML applications involving physics-driven datasets. Neural operators are grid-resolution-invariant and are often evaluated on higher grid resolutions than they are trained on, i.e., zero-shot super-resolution. Given their promising zero-shot super-resolution performance on dynamical systems emulation, we present a critical investigation of their zero-shot weather downscaling capabilities, which is when models are tasked with producing high-resolution outputs using higher upsampling factors than are seen during training. To this end, we create two realistic downscaling experiments with challenging upsampling factors (e.g., 8x and 15x) across data from different simulations: the European Centre for Medium-Range Weather Forecasts Reanalysis version 5 (ERA5) and the Wind Integration National Dataset Toolkit. While neural operator-based downscaling models perform better than interpolation and a simple convolutional baseline, we show the surprising performance of an approach that combines a powerful transformer-based model with parameter-free interpolation at zero-shot weather downscaling. We find that this Swin-Transformer-based approach mostly outperforms models with neural operator layers in terms of average error metrics, whereas an Enhanced Super-Resolution Generative Adversarial Network-based approach is better than most models in terms of capturing the physics of the ground truth data. We suggest their use in future work as strong baselines.
Zitieren
APA
Sinha, S., Benton, B., & Emami, P. (2025). On the effectiveness of neural operators at zero-shot weather downscaling. Environmental Data Science, 4. https://doi.org/10.1017/eds.2025.11
Vancouver
Sinha S, Benton B, Emami P. On the effectiveness of neural operators at zero-shot weather downscaling. Environmental Data Science. 2025;4. doi:10.1017/eds.2025.11
BibTeX
@article{Sinha2025,
author = {Sinha, Saumya and Benton, Brandon and Emami, Patrick},
title = {On the effectiveness of neural operators at zero-shot weather downscaling},
journal = {Environmental Data Science},
year = {2025},
volume = {4},
doi = {10.1017/eds.2025.11}
}
EUVIMED