Synthetic Aperture Radar (SAR) is a unique remote sensing instrument imaging ocean surface waves in two dimensions with high spatial resolution regardless of sunlight and weather conditions. However, due to the nonlinear imaging process, the ocean wave spectra cannot be retrieved directly from SAR data. The emergence of deep learning (DL) techniques provides a new paradigm for addressing this challenge. In this paper, a deep-learning-based model, called Wave-Spec-CNN, is proposed for retrieving omni-directional ocean wave spectra from SAR data. This model is constructed using approximately 21,000 collocations of Sentinel-1 Interferometric Wide swath mode images matched with global in-situ buoy data. The model adapts the convolution neural network (CNN) to accommodate the multi-valued nature of omni-directional ocean wave spectra, enhances performance by integrating a calibration branch and further incorporates physical characteristics into the training process. The results demonstrate consistency with buoy measurements for significant wave height (SWH) in the range of 0.5 m to 6 m, yielding a root-mean-square error (RMSE) of 0.51 m on the validation dataset, comparable to traditional physical-based methods. In terms of mean wave period (MWP) and peak frequency (PF), the achieved RMSEs are of 1.24 s and 0.03 Hz, respectively. The retrieved omni-directional ocean wave spectra also allow to separate swell and windsea components for respective comparisons with those derived by in-situ buoy data. The RMSEs of respective SWH comparisons are of 0.46 m and 0.42 m. This research represents an initial endeavor into utilizing DL for the long-standing challenge of SAR inversion for ocean wave spectra, as well as providing valuable insights for employing DL in multi-parameter inversion tasks in remote sensing.

中文翻译：

---

从星载合成孔径雷达中检索全向海波谱的深度学习


合成孔径雷达（SAR）是一种独特的遥感仪器，无论阳光和天气条件如何，都能以高空间分辨率对海洋表面波浪进行二维成像。然而，由于非线性成像过程，无法直接从SAR数据中反演海浪谱。深度学习（DL）技术的出现为应对这一挑战提供了新的范例。本文提出了一种名为 Wave-Spec-CNN 的基于深度学习的模型，用于从 SAR 数据中检索全向海浪频谱。该模型是使用大约 21,000 个 Sentinel-1 干涉宽幅模式图像与全球现场浮标数据匹配的搭配构建的。该模型采用卷积神经网络（CNN）来适应全向海浪频谱的多值性质，通过集成校准分支来增强性能，并进一步将物理特征纳入训练过程。结果表明，与 0.5 m 至 6 m 范围内的有义波高 (SWH) 的浮标测量结果一致，验证数据集上的均方根误差 (RMSE) 为 0.51 m，与传统的基于物理的方法相当。就平均波周期 (MWP) 和峰值频率 (PF) 而言，实现的 RMSE 分别为 1.24 s 和 0.03 Hz。检索到的全向海浪谱还可以分离海浪和风海分量，以便分别与现场浮标数据得出的数据进行比较。各个 SWH 比较的 RMSE 分别为 0.46 m 和 0.42 m。 这项研究代表了利用深度学习解决海浪频谱SAR反演长期挑战的初步努力，并为在遥感多参数反演任务中使用深度学习提供了宝贵的见解。