Gespeichert in:
| Hauptverfasser: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | https://arxiv.org/abs/2602.11807 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866912901697110016 |
|---|---|
| author | Wu, Lianjun Zhu, Shengchen Liu, Yuxuan Kai, Liuyu Feng, Xiaoduan Wang, Duomin Liu, Wenshuo Zhang, Jingxuan Li, Kelvin Wang, Bin |
| author_facet | Wu, Lianjun Zhu, Shengchen Liu, Yuxuan Kai, Liuyu Feng, Xiaoduan Wang, Duomin Liu, Wenshuo Zhang, Jingxuan Li, Kelvin Wang, Bin |
| contents | Latent diffusion models (LDMs) suffer from limited diffusability in high-resolution (<=0.25°) ensemble weather forecasting, where diffusability characterizes how easily a latent data distribution can be modeled by a diffusion process. Unlike natural image fields, meteorological fields lack task-agnostic foundation models and explicit semantic structures, making VFM-based regularization inapplicable. Moreover, existing frequency-based approaches impose identical spectral regularization across channels under a homogeneity assumption, which leads to uneven regularization strength under the inter-variable spectral heterogeneity in multivariate meteorological data. To address these challenges, we propose a 3D Masked AutoEncoder (3D-MAE) that encodes weather-state evolution features as an additional conditioning for the diffusion model, together with a Variable-Aware Masked Frequency Modeling (VA-MFM) strategy that adaptively selects thresholds based on the spectral energy distribution of each variable. Together, we propose PuYun-LDM, which enhances latent diffusability and achieves superior performance to ENS at short lead times while remaining comparable to ENS at longer horizons. PuYun-LDM generates a 15-day global forecast with a 6-hour temporal resolution in five minutes on a single NVIDIA H200 GPU, while ensemble forecasts can be efficiently produced in parallel. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_11807 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | PuYun-LDM: A Latent Diffusion Model for High-Resolution Ensemble Weather Forecasts Wu, Lianjun Zhu, Shengchen Liu, Yuxuan Kai, Liuyu Feng, Xiaoduan Wang, Duomin Liu, Wenshuo Zhang, Jingxuan Li, Kelvin Wang, Bin Artificial Intelligence Latent diffusion models (LDMs) suffer from limited diffusability in high-resolution (<=0.25°) ensemble weather forecasting, where diffusability characterizes how easily a latent data distribution can be modeled by a diffusion process. Unlike natural image fields, meteorological fields lack task-agnostic foundation models and explicit semantic structures, making VFM-based regularization inapplicable. Moreover, existing frequency-based approaches impose identical spectral regularization across channels under a homogeneity assumption, which leads to uneven regularization strength under the inter-variable spectral heterogeneity in multivariate meteorological data. To address these challenges, we propose a 3D Masked AutoEncoder (3D-MAE) that encodes weather-state evolution features as an additional conditioning for the diffusion model, together with a Variable-Aware Masked Frequency Modeling (VA-MFM) strategy that adaptively selects thresholds based on the spectral energy distribution of each variable. Together, we propose PuYun-LDM, which enhances latent diffusability and achieves superior performance to ENS at short lead times while remaining comparable to ENS at longer horizons. PuYun-LDM generates a 15-day global forecast with a 6-hour temporal resolution in five minutes on a single NVIDIA H200 GPU, while ensemble forecasts can be efficiently produced in parallel. |
| title | PuYun-LDM: A Latent Diffusion Model for High-Resolution Ensemble Weather Forecasts |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2602.11807 |