Topography-Enhanced Multilevel Residual Attention U-Net Model for Sea Ice Concentration Spatial Super-Resolution Prediction
High-resolution (HR) sea ice concentration (SIC) is essential for polar ocean monitoring and Arctic research. Super-resolution (SR) techniques improve the spatial resolution of SIC data, but existing methods often fail to capture fine-scale structures, perform poorly at ice edges, and lack accuracy...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11106744/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | High-resolution (HR) sea ice concentration (SIC) is essential for polar ocean monitoring and Arctic research. Super-resolution (SR) techniques improve the spatial resolution of SIC data, but existing methods often fail to capture fine-scale structures, perform poorly at ice edges, and lack accuracy near sea–land boundaries. To address these challenges, we propose a topography-enhanced multilevel residual attention U-Net (TE-MRAU-Net) for SIC downscaling. To address these challenges, this article proposes a TE-MRAU-Net downscaling model. TE-MRAU-Net integrates three innovative modules: the HR topography feature module, which introduces static topographic constraints to effectively improve reconstruction accuracy along sea–land boundaries; the multilevel residual module, which enhances the model’s ability to extract fine-scale sea ice features in super-resolution predictions; and the spatial attention connector module, which strengthens spatial modeling and structural consistency, particularly improving reconstruction performance in marginal sea ice edges and lower latitude Arctic regions. Experimental results indicate that, in the pan-Arctic region, TE-MRAU-Net outperforms other models in various evaluation metrics and spatial distribution reconstruction. For the 5× SR prediction task, the model achieves a mean absolute error of 0.0049, a root-mean-square error of 0.0081, a correlation coefficient (<inline-formula><tex-math notation="LaTeX">$r$</tex-math></inline-formula>) of 0.9993, a peak signal-to-noise ratio of 40.21 dB, a structural similarity index measure of 0.9764, a balanced accuracy of 0.9428, a mean opinion score of 4.63, and a subjective image quality assessment of 9.75, greatly enhancing the model’s ability to capture detailed SIC features. |
|---|---|
| ISSN: | 1939-1404 2151-1535 |