SCAWaveNet: A Spatial-Channel Attention-based Network for Global Significant Wave Height Retrieval

2025-07-01

AI Summary

To address insufficient cross-channel information interaction in four-channel Delay-Doppler Map (DDM) data from the Cyclone Global Navigation Satellite System (CYGNSS) for significant wave height (SWH) retrieval, this paper proposes SCAWaveNet—a lightweight neural network integrating dual-dimensional spatial and channel attention mechanisms. Innovatively incorporating a multi-head attention module, SCAWaveNet models nonlinear interdependencies among multi-channel DDM features, enabling adaptive joint spatial-channel weighting and fusion. Compared with conventional single-channel or naive concatenation approaches, SCAWaveNet substantially enhances remote sensing feature representation. Evaluated against ERA5 reanalysis and NDBC buoy reference datasets, it achieves SWH retrieval RMSEs of 0.438 m and 0.432 m, respectively—improving upon the current state-of-the-art by 3.52% (ERA5) and 5.47% (NDBC). These results validate the effectiveness of cross-channel attention modeling for GNSS-R ocean remote sensing retrieval.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Data management and stream processing for Web, mobile and wireless applications

Abstract

Recent advancements in spaceborne GNSS missions have produced extensive global datasets, providing a robust basis for deep learning-based significant wave height (SWH) retrieval. While existing deep learning models predominantly utilize CYGNSS data with four-channel information, they often adopt single-channel inputs or simple channel concatenation without leveraging the benefits of cross-channel information interaction during training. To address this limitation, a novel spatial-channel attention-based network, namely SCAWaveNet, is proposed for SWH retrieval. Specifically, features from each channel of the DDMs are modeled as independent attention heads, enabling the fusion of spatial and channel-wise information. For auxiliary parameters, a lightweight attention mechanism is designed to assign weights along the spatial and channel dimensions. The final feature integrates both spatial and channel-level characteristics. Model performance is evaluated using four-channel CYGNSS data. When ERA5 is used as a reference, SCAWaveNet achieves an average RMSE of 0.438 m. When using buoy data from NDBC, the average RMSE reaches 0.432 m. Compared to state-of-the-art models, SCAWaveNet reduces the average RMSE by at least 3.52% on the ERA5 dataset and by 5.47% on the NDBC buoy observations. The code is available at https://github.com/Clifx9908/SCAWaveNet.

Problem

• Improves significant wave height retrieval using spatial-channel attention
• Enhances cross-channel information interaction in deep learning models
• Reduces RMSE compared to state-of-the-art SWH retrieval methods

Innovation

• Spatial-channel attention network for SWH retrieval
• Multi-channel feature fusion via attention heads
• Lightweight attention for auxiliary parameter weighting