Bo et al. (2026) Physically Consistent Reconstruction of Sparse Scatterometer Ocean Surface Wind Fields Based on Physics‐Guided Generative Learning

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Geophysical Research Letters
• Year: 2026
• Date: 2026-08-13
• Authors: Ran Bo, Zengliang Zang, Zeming Zhou, Pinglv Yang, Huadong Du, Xiaofeng Zhao, Qian Li, Yangfan Hu
• DOI: 10.1029/2026gl123802

Research Groups

Not specified in the provided text.

Short Summary

The study develops a physics-guided generative learning network to reconstruct sea surface wind vector fields from sparse scatterometer data, ensuring physical consistency and high accuracy.

Objective

• To reconstruct accurate and timely sea surface wind fields from sparse scatterometer-derived observations by integrating data-driven learning with physics-based constraints.

Study Configuration

• Spatial Scale: Sea surface (regional/oceanic).
• Temporal Scale: Includes a case study of Typhoon Merbok in September 2022.

Methodology and Data

• Models used: Physics-guided generative learning network (comprising a generator with a composite physical loss function and a discriminator utilizing physics consistency scores).
• Data sources: Scatterometer-derived observations.

Main Results

• The proposed network produces fast, structurally preservative, and physically consistent wind field reconstructions.
• During the analysis of Typhoon Merbok (September 2022), the model successfully recovered the vortex structure with a wind speed Root Mean Square Error (RMSE) of $\le 2\text{ m/s}$.
• The method demonstrates robustness in handling extreme weather events even under conditions of significant data sparsity.

Contributions

• Introduces a novel framework that combines generative adversarial learning with physics-based guiding terms to overcome the limitations of purely data-driven reconstructions in oceanography.

Funding

Not specified in the provided text.

Citation

@article{Bo2026Physically,
  author = {Bo, Ran and Zang, Zengliang and Zhou, Zeming and Yang, Pinglv and Du, Huadong and Zhao, Xiaofeng and Li, Qian and Hu, Yangfan},
  title = {Physically Consistent Reconstruction of Sparse Scatterometer Ocean Surface Wind Fields Based on Physics‐Guided Generative Learning},
  journal = {Geophysical Research Letters},
  year = {2026},
  doi = {10.1029/2026gl123802},
  url = {https://doi.org/10.1029/2026gl123802}
}