Satellite remote sensing plays a fundamental role in observing oceanic processes by providing large-scale, long-term, and continuous measurements. With the increasing availability of multisource satellite data, challenges such as data gaps, complex environmental conditions, and the limitations of conventional retrieval methods have become more evident. In recent years, artificial intelligence (AI) has emerged as a practical and effective approach to address these issues. This article reviews the development of AI techniques in satellite ocean remote sensing, focusing on three main application areas: parameter retrieval, data reconstruction, and image-based ocean phenomenon detection. For geophysical variable retrieval, AI models such as convolutional neural networks (CNNs) and Transformer architectures have improved the accuracy of ocean waves, sea surface, salinity, wind, and ocean color estimates, especially under extreme or noisy conditions. In the field of data reconstruction, AI methods enable the completion of missing data in both surface and subsurface ocean layers, offering finer spatial–temporal resolution and better consistency than traditional interpolation approaches. For image interpretation, deep learning (DL) models have been applied to detect and segment dynamic ocean features such as mesoscale eddies, internal waves, sea ice, and tropical cyclones (TCs), achieving high efficiency and precision. This article also highlights the integration of AI with physical knowledge, the use of multisource fusion, and the trend toward near real-time (NRT) applications. These developments indicate that AI will play an increasingly important role in future satellite-based ocean observation and environmental monitoring.

This article introduces an innovative deep-learning approach for retrieving tropical cyclone (TC) rainfall information from C-band Sentinel-1 synthetic aperture radar (SAR) imagery. We collected 17 SAR images under TC conditions from 2016 to 2021 and matched them with synchronous observational Next Generation Weather Radar (NEXRAD) Level-III data, forming a dataset of 302689 data pairs for model development. The model inputs include SAR-measured physical parameters in normalized radar cross section (NRCS), texture features represented by the gray-level co-occurrence matrix (GLCM), and statistical parameters of VV-polarized NRCS. A deep-learning-based TC rain rate retrieval (TC3R) model, combining a convolutional network and a fully connected (FC) network, was developed to retrieve quantitative TC rainfall information effectively. The test results demonstrate that the TC3R model can offer reasonable and stable quantitative rainfall estimation, particularly effectively detecting areas with medium-to-heavy rainfall events (2.5–40 mm/h) in SAR images where the NRCS is significantly affected by rain. Furthermore, to offer valuable insights into the performance of the TC3R model, we analyzed results across TC events of different intensities as case studies. Our results show high structural similarity (SSIM) in rainfall patterns between SAR and NEXRAD across all cases, consistently achieving SSIM values above 0.67. Moreover, in areas where SAR signals are notably affected by rainfall, the SSIM index even exceeds 0.80. Finally, our model’s performance was evaluated by comparing its results with the independent global precipitation measurement (GPM) data, demonstrating effective rainfall prediction, particularly for the primary spiral rain band, in the two cases analyzed.

This study developed a machine-learning-based model to extract the rainfall information from C-band synthetic aperture radar (SAR) images acquired in dual-polarization (VV/VH) over hurricane conditions. The model is based on a back-propagation neural network (BPNN) tuned by 1,1762 pairs of samples from collocations of Sentinel-1 data and Stepped Frequency Microwave Radiometer (SFMR) measurements. The model inputs include four SAR measured physical parameters and one morphological feature related to the hurricane. Several comparative tests show that the selection of different SAR inputs has a significant impact on the performance of the models for rainfall estimation. For example, compared to the SFMR measurements, the root mean square error and correlation coefficient of rain rate obtained by BPNN with all five inputs reach the best, which are 3.61 mm/hr and 0.87, respectively.

This study developed a deep-learning-based model to retrieve sea surface hurricane winds from synthetic aperture radar (SAR) imagery. We introduce the essential idea, residual learning, of the Residual Net into the artificial neural network and design a deep cross-layer concatenation network. The model inputs include SAR measured physical parameters in backscattering energy, the texture feature represented by the grey level co-occurrence matrix, and the morphological hurricane feature. We collected 45 satellite SAR images from Sentinel-1 over hurricane conditions. These images were divided into 39 and 6 for model development and independent testing. A total of 16,127 wind samples acquired from 39 SAR images and simultaneously measured by the Stepped Frequency Microwave Radiometer were collected as model tuning datasets, among which 80% and 20% were used for training and validation. Our validation results show that the deep-learning-based model achieved a correlation coefficient (CORR) and root-mean-square error (RMSE) of 0.98 and 1.72 m/s for wind speeds up to 75 m/s. We further applied the model to six independent SAR images. The model significantly outperformed two existing geophysical algorithms and one backpropagation neural network algorithm with the RMSE is 2.61 m/s and a CORR of 0.95. Moreover, statistical analysis in different wind speed regimes indicates that our model shows a stable performance improvement than comparable algorithms. The RMSE decreases 10%~ 70%, especially the reduction of RMSE is more than 45% at high wind speed (> 42 m/s). Furthermore, adding an independent rainfall estimate to the deep-learning model further enhanced the wind retrieval algorithm.

With the increase of the amount of space-borne synthetic aperture radar (SAR), SAR data is increasingly being applied for remote monitoring of hurricane wind speed filed. There are many geophysical model functions (GMFs) have been developed for wind speed inversion by describing the relationship between the surface winds and the normalized radar backscatter cross section (NRCS) of SAR. In this paper, we provide a method based on support vector machine (SVM) for retrieving oceanic surface wind speeds over hurricanes. But unlike most traditional GMFs, this method does not need formula fitting for the input parameters which include dual polarization (VV+VH) SAR normalized radar cross section and incidence angle (θ). In addition, we use the Stepped Frequency Microwave Radiometer (SFMR) wind filed data as the training and validation data for SVM model. The results show that the SVM model has achieved good results over totally independent testing data. The retrieved hurricane wind speed are in good agreement with the SFMR wind filed, and the correlation coefficient and root mean square error(RMSE) were 0.91 and 4.20 m/s, respectively.