Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2505.10894 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910947627499520 |
|---|---|
| author | Wang, Yishuo Zhou, Feng Zhou, Muping Meng, Qicheng Hu, Zhijun Wang, Yi |
| author_facet | Wang, Yishuo Zhou, Feng Zhou, Muping Meng, Qicheng Hu, Zhijun Wang, Yi |
| contents | This paper proposes CTP, a novel deep learning framework that integrates convolutional neural network(CNN), Transformer architectures, and physics-informed neural network(PINN) for ocean front prediction. Ocean fronts, as dynamic interfaces between distinct water masses, play critical roles in marine biogeochemical and physical processes. Existing methods such as LSTM, ConvLSTM, and AttentionConv often struggle to maintain spatial continuity and physical consistency over multi-step forecasts. CTP addresses these challenges by combining localized spatial encoding, long-range temporal attention, and physical constraint enforcement. Experimental results across south China sea(SCS) and Kuroshio(KUR) regions from 1993 to 2020 demonstrate that CTP achieves state-of-the-art(SOTA) performance in both single-step and multi-step predictions, significantly outperforming baseline models in accuracy, $F_1$ score, and temporal stability. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_10894 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | CTP: A hybrid CNN-Transformer-PINN model for ocean front forecasting Wang, Yishuo Zhou, Feng Zhou, Muping Meng, Qicheng Hu, Zhijun Wang, Yi Machine Learning Computer Vision and Pattern Recognition This paper proposes CTP, a novel deep learning framework that integrates convolutional neural network(CNN), Transformer architectures, and physics-informed neural network(PINN) for ocean front prediction. Ocean fronts, as dynamic interfaces between distinct water masses, play critical roles in marine biogeochemical and physical processes. Existing methods such as LSTM, ConvLSTM, and AttentionConv often struggle to maintain spatial continuity and physical consistency over multi-step forecasts. CTP addresses these challenges by combining localized spatial encoding, long-range temporal attention, and physical constraint enforcement. Experimental results across south China sea(SCS) and Kuroshio(KUR) regions from 1993 to 2020 demonstrate that CTP achieves state-of-the-art(SOTA) performance in both single-step and multi-step predictions, significantly outperforming baseline models in accuracy, $F_1$ score, and temporal stability. |
| title | CTP: A hybrid CNN-Transformer-PINN model for ocean front forecasting |
| topic | Machine Learning Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2505.10894 |