Saved in:
| Main Authors: | , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2507.09202 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908818752929792 |
|---|---|
| author | Wang, Wuxin Ni, Weicheng Huang, Lilan Hao, Tao Fei, Ben Ma, Shuo Yuan, Taikang Zhao, Yanlai Deng, Kefeng Li, Xiaoyong Leng, Hongze Duan, Boheng Bai, Lei Zhang, Weimin Song, Junqiang Ren, Kaijun |
| author_facet | Wang, Wuxin Ni, Weicheng Huang, Lilan Hao, Tao Fei, Ben Ma, Shuo Yuan, Taikang Zhao, Yanlai Deng, Kefeng Li, Xiaoyong Leng, Hongze Duan, Boheng Bai, Lei Zhang, Weimin Song, Junqiang Ren, Kaijun |
| contents | Machine Learning (ML) has shown great promise in revolutionizing weather forecasting, yet most ML systems still rely on initial conditions generated by Numerical Weather Prediction (NWP) systems. End-to-end ML models aim to eliminate this dependency, but they often rely on observation-specific encoders and require redesign or retraining when observation sources change, thereby limiting their operational robustness. Here, we introduce XiChen, a global weather observation-to-forecast ML system via four-dimensional variational (4DVar) gradient-guided flexible assimilation. We demonstrate that the gradient of the 4DVar cost function serves as a physically grounded interface that maps heterogeneous observations into a common state space. This novel formulation enables XiChen to flexibly assimilate diverse conventional and raw satellite observations while preserving physical consistency. Experiments show that the system achieves forecasting metrics competitive with operational NWP systems. This work provides a practical and physically consistent route toward operational ML-based global weather forecasting systems with heterogeneous and evolving observations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_09202 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | XiChen: A global weather observation-to-forecast machine learning system via four-dimensional variational gradient-guided flexible assimilation Wang, Wuxin Ni, Weicheng Huang, Lilan Hao, Tao Fei, Ben Ma, Shuo Yuan, Taikang Zhao, Yanlai Deng, Kefeng Li, Xiaoyong Leng, Hongze Duan, Boheng Bai, Lei Zhang, Weimin Song, Junqiang Ren, Kaijun Machine Learning Artificial Intelligence Atmospheric and Oceanic Physics Machine Learning (ML) has shown great promise in revolutionizing weather forecasting, yet most ML systems still rely on initial conditions generated by Numerical Weather Prediction (NWP) systems. End-to-end ML models aim to eliminate this dependency, but they often rely on observation-specific encoders and require redesign or retraining when observation sources change, thereby limiting their operational robustness. Here, we introduce XiChen, a global weather observation-to-forecast ML system via four-dimensional variational (4DVar) gradient-guided flexible assimilation. We demonstrate that the gradient of the 4DVar cost function serves as a physically grounded interface that maps heterogeneous observations into a common state space. This novel formulation enables XiChen to flexibly assimilate diverse conventional and raw satellite observations while preserving physical consistency. Experiments show that the system achieves forecasting metrics competitive with operational NWP systems. This work provides a practical and physically consistent route toward operational ML-based global weather forecasting systems with heterogeneous and evolving observations. |
| title | XiChen: A global weather observation-to-forecast machine learning system via four-dimensional variational gradient-guided flexible assimilation |
| topic | Machine Learning Artificial Intelligence Atmospheric and Oceanic Physics |
| url | https://arxiv.org/abs/2507.09202 |