AGCD: Agent-Guided Cross-Modal Decoding for Weather Forecasting
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866917347156033536 |
|---|---|
| author | Wu, Jing Liu, Yang Zhang, Lin Zeng, Junbo Wang, Jiabin Ye, Zi Li, Guowen Cao, Shilei Cheng, Jiashun Wang, Fang Jin, Meng Feng, Yerong Cheng, Hong Lu, Yutong Fu, Haohuan Zheng, Juepeng |
| author_facet | Wu, Jing Liu, Yang Zhang, Lin Zeng, Junbo Wang, Jiabin Ye, Zi Li, Guowen Cao, Shilei Cheng, Jiashun Wang, Fang Jin, Meng Feng, Yerong Cheng, Hong Lu, Yutong Fu, Haohuan Zheng, Juepeng |
| contents | Accurate weather forecasting is more than grid-wise regression: it must preserve coherent synoptic structures and physical consistency of meteorological fields, especially under autoregressive rollouts where small one-step errors can amplify into structural bias. Existing physics-priors approaches typically impose global, once-for-all constraints via architectures, regularization, or NWP coupling, offering limited state-adaptive and sample-specific controllability at deployment. To bridge this gap, we propose Agent-Guided Cross-modal Decoding (AGCD), a plug-and-play decoding-time prior-injection paradigm that derives state-conditioned physics-priors from the current multivariate atmosphere and injects them into forecasters in a controllable and reusable way. Specifically, We design a multi-agent meteorological narration pipeline to generate state-conditioned physics-priors, utilizing MLLMs to extract various meteorological elements effectively. To effectively apply the priors, AGCD further introduce cross-modal region interaction decoding that performs region-aware multi-scale tokenization and efficient physics-priors injection to refine visual features without changing the backbone interface. Experiments on WeatherBench demonstrate consistent gains for 6-hour forecasting across two resolutions (5.625 degree and 1.40625 degree) and diverse backbones (generic and weather-specialized), including strictly causal 48-hour autoregressive rollouts that reduce early-stage error accumulation and improve long-horizon stability. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_15260 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | AGCD: Agent-Guided Cross-Modal Decoding for Weather Forecasting Wu, Jing Liu, Yang Zhang, Lin Zeng, Junbo Wang, Jiabin Ye, Zi Li, Guowen Cao, Shilei Cheng, Jiashun Wang, Fang Jin, Meng Feng, Yerong Cheng, Hong Lu, Yutong Fu, Haohuan Zheng, Juepeng Artificial Intelligence Computer Vision and Pattern Recognition Accurate weather forecasting is more than grid-wise regression: it must preserve coherent synoptic structures and physical consistency of meteorological fields, especially under autoregressive rollouts where small one-step errors can amplify into structural bias. Existing physics-priors approaches typically impose global, once-for-all constraints via architectures, regularization, or NWP coupling, offering limited state-adaptive and sample-specific controllability at deployment. To bridge this gap, we propose Agent-Guided Cross-modal Decoding (AGCD), a plug-and-play decoding-time prior-injection paradigm that derives state-conditioned physics-priors from the current multivariate atmosphere and injects them into forecasters in a controllable and reusable way. Specifically, We design a multi-agent meteorological narration pipeline to generate state-conditioned physics-priors, utilizing MLLMs to extract various meteorological elements effectively. To effectively apply the priors, AGCD further introduce cross-modal region interaction decoding that performs region-aware multi-scale tokenization and efficient physics-priors injection to refine visual features without changing the backbone interface. Experiments on WeatherBench demonstrate consistent gains for 6-hour forecasting across two resolutions (5.625 degree and 1.40625 degree) and diverse backbones (generic and weather-specialized), including strictly causal 48-hour autoregressive rollouts that reduce early-stage error accumulation and improve long-horizon stability. |
| title | AGCD: Agent-Guided Cross-Modal Decoding for Weather Forecasting |
| topic | Artificial Intelligence Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2603.15260 |