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.

Green macroalgae (GMA) are prominently visible as distinctive bright mats on Synthetic Aperture Radar (SAR) imagery that are frequently used to complement conventional optical observations for extracting GMA distribution at sea. However, the mechanisms ruling the interaction between the microwaves and the GMA are not well understood. This study aims to address this gap by systematically analyzing C-band backscattering from GMA-covered sea surfaces from a physical perspective using dual-polarized Sentinel-1 SAR imagery combined with a three-layer description of the GMA-covered sea surface. Firstly, the dual-polarized normalized radar cross-section (NRCS) is analyzed showing that the co-polarized backscatter is always above the system noise while this is not the case for the cross-polarized backscatter that is frequently noisy. Then, polarimetric descriptors are adopted to shed light on the GMA backscatter mechanisms. The degree of polarization (DoP) is evaluated to demonstrate that the electromagnetic (EM) wave backscattered from the GMA is almost fully polarized, indicating a negligible depolarized component. Finally, the polarization ellipse associated to the EM wave backscattered from the GMA is analyzed showing the GMA call for a linearly-polarized (almost vertically oriented) backscattered EM wave. This finding suggests a scattering mechanism dominated by single reflection and a residual multiple reflection component within the wet-GMA layer of the three-layer schematic model. The analysis is extended to several imagery collected under different green tide stages showing that the above-mentioned findings always apply.

With the development of deep learning, impressive progress has been made in the field of image restoration. The existing methods mainly rely on CNN and Transformer to obtain multi-scale feature information. However, these methods rarely integrate frequency domain information effectively during feature extraction, limiting their performance in image restoration. Additionally, few have combined Mamba with the Fourier domain for image restoration, which limits Mamba’s ability to perceive global degradation in the frequency domain. Therefore, we propose a new image restoration model called FMambaIR, which utilizes the complementarity between frequency and Mamba for image restoration. The core of FMambaIR is the F-Mamba block, which combines Fourier transform and Mamba for global degradation perception modeling. Specifically, F-Mamba adopts a dual branch complementary structure, including spatial Mamba branches and Fourier frequency domain global modeling. Mamba models the long-range dependencies of the entire image features, and the frequency branch utilizes Fourier to extract global degraded features from the image. Finally, we use a forward feedback network to integrate local information, which is beneficial for improving the recovery details. We comprehensively evaluate FMambaIR on several image restoration tasks, including underwater image enhancement, remote sensing image dehazing, and low-light image enhancement. The experimental results demonstrate that FMambaIR not only achieves superior performance compared to state-of-the-art methods but also significantly reduces computational complexity. Our code is available at https://github.com/mickoluan/FMambaIR.

Synthetic aperture radar (SAR) is widely recognized as an effective tool for investigating oceanic processes, particularly internal wave dynamics, due to its ability to capture high-resolution imaging across diverse environmental conditions. However, speckle noise in SAR imagery poses significant challenges for accurately extracting critical parameters, such as wavefront characteristics and bright-dark distances, thereby hindering its application in detailed internal wave studies. To address this, a novel speckle filtering method is proposed, specifically designed to enhance the quality of SAR images capturing the feature of internal wave. First, a structural description matrix with complex values is introduced to encode pixel value correlations and local spatial features, enabling improved representation of wave-related contextual information within the SAR data. Second, a similarity testing framework based on the Wishart distribution is developed to identify and aggregate structurally similar regions, ensuring robust noise suppression while preserving essential structural parameters. The proposed method strikes a balance between speckle noise reduction and structural preservation, making it adaptable to a variety of SAR datasets. Experimental validation on SAR imagery of oceanic internal waves demonstrates the method’s effectiveness in retaining key internal wave information, while additional tests on terrestrial targets highlight its potential for generalization across different scenarios.

Air-sea heat fluxes play a crucial role in understanding global climate variability. Using the bulk aerodynamic algorithm, we can derive sensible heat flux (SHF) and latent heat flux (LHF) from the satellite sea surface temperature (Ts) and wind speed (WS), as well as the air temperature (Ta) and specific humidity (Qa). However, traditional retrievals ofTaandQatend to be unreliable. To address this issue, we introduced the Matrices-Points Fusion Network (MPFNet), a deep learning (DL) model designed to integrate spatial and point information. This model employs the Fourier Neural Operator (FNO) and Residual Network (ResNet) techniques to enhance retrieval capabilities. The model was pre-trained with ECMWF ERA5 reanalysis-based data, followed by transfer learning (TL) with satellite and in situ matchup data for fine-tuning. Validation against independent in situ data showed significant improvements compared to the mainstream products in the community, with root mean square errors (RMSE) for Taand Qareduced to 0.59°C and 0.87g/kg, representing 27% to 41% and 16% to 33% improvements, respectively.SHFandLHFRMSE values were 6.54 W/m2and 29.32 W/m2, reflecting improvements of 32% to 36% and 17% to 31%, respectively. Using this fine-tuned MPFNet model with satellite data as input, a global daily gridded dataset of Ta,Qa,SHF, andLHFover 11.5 years was generated at a 0.25° resolution. The analysis showed that previous products tended to underestimate Taand Qawhile overestimatingSHFandLHF.

Ocean color remote sensing has advanced over the past 40 years, with algorithms primarily rooted in radiative transfer theory to estimate various ocean color properties. To address the diversity and region-specific distribution of these properties, numerous algorithms have been developed; however, each is tailored to specific variables. In recent years, deep learning has demonstrated remarkable progress in this field, but a general foundational model for robust ocean color product generation remains lacking. To address this limitation, this study introduces the Ocean Color deep learning Foundation Model (OCFM), a comprehensive framework designed to extract multiple global ocean color properties from satellite observations. The development of OCFM follows three sequential phases: pre-training with operational satellite products to learn fundamental theoretical relationships, fine-tuning with in-situ measurements to align with real ocean state, and deploying the trained OCFM for flexible retrieval of ocean color properties. Quantitative evaluation shows that, in end-user application tests, the model achieved a coefficient of determination (R2) of 0.90 for primary productivity and 0.71 for water clarity, with corresponding mean absolute percentage differences of 44.08% and 17.28%, respectively. OCFM provides a more efficient solution for few-shot and unevenly distributed samples compared to traditional retrieval algorithms. Even with limited user resources and small sample sizes, additional downstream training with the trained OCFM can achieve state-of-the-art performance in retrieving ocean color properties. This study highlights the potential of a deep learning foundation model for generalizable and few-shot ocean color retrieval.

The Kuroshio intrusion into the South China Sea via the Luzon Strait significantly influences regional ocean dynamics. However, predicting this intrusion, especially the Kuroshio Loop Current (KLC), remains challenging due to its complex mesoscale and submesoscale processes. Traditional physical models struggle to capture the nonlinear and multiscale features of the Kuroshio intrusion, while deep learning approaches face challenges in incorporating the essential physical processes that characterize the KLC. To address these challenges, we developed the Kuroshio Intrusion Forecast Network (KIFnet), a multi-prior knowledge guided deep learning model. KIFnet integrates physical oceanographic principles with data-driven predictions, enhancing its ability to capture complex ocean dynamics. KIFnet incorporates an SST-guided SSH prediction module and a vorticity-guided loss function to explicitly model thermal and dynamic features of the KLC, advancing the challenging task of forecasting KLC intrusion events. Experimental results demonstrate the model achieves an accuracy of 88% for KLC intrusion events and provides reliable predictions up to 10 days ahead. Prior limitations in KLC forecasting have constrained SCS climate modeling and marine ecosystem management. KIFnet provides accurate KLC predictions, supporting proactive climate adaptation and sustainable ecosystem strategies.

This study proposed an improved Doppler radar imaging model (IDopRIM) to address the overestimation of the original DopRIM and produce better accuracy for investigating the current-wave interaction effect on ocean surface current (OSC) retrieval. The IDopRIM’s performance, with the root-mean-square (rms) errors mostly under 0.3 m/s, is validated against the empirical model and actual measurements. It shows a notable improvement over the original model in which the maximum deviation could be up to about 1.5 m/s at moderate-to-high wind speeds and upwind direction for HH polarization. Furthermore, this study uses both numerical simulations and real synthetic aperture radar (SAR) imagery to assess the impact of current-wave interaction on OSC retrieval, especially in scenarios involving ocean internal waves (IWs), using the Doppler centroid anomaly (DCA) method incorporated with the proposed IDopRIM. The results underscore the significance of incorporating current-wave interactions in OSC retrieval for IW conditions, revealing that neglecting these interactions can result in relative errors of over 30% in certain cases.

The detection and preservation of marine biodiversity has garnered global attention. The incorporation of deep learning methodologies can elevate the efficiency of species detection. In this study, we developed a DeepSeaNet for effective localization and accurate classification of organisms based on deep-sea images, as well as for hinting at unknown organisms (new species). The DeepSeaNet fully accommodates the unique characteristics of deep-sea organisms and imaging environment, leading to remarkable advancements in fine-grained analysis and accuracy. The DeepSeaNet comprises two network components: a deep-sea Classes Detection Network (CDN) and an unsupervised Species Clustering Network (SCN). CDN is used for biological class detection and is specifically tailored for deep-sea environments. It incorporates modules for feature fusion, multi-scale analysis, and self-attention. SCN is specifically designed to detect and identify new species by utilizing the location information extracted from the CDN output results. It is composed of a feature extraction module and a clustering module. By collecting deep-sea image data from the “KeXue” Science Research Vessel, we constructed a dataset totaling 29,436 images of deep-sea organisms covering more than 500 species of deep-sea seamount organisms. This dataset serves as the foundational dataset for our experiment. As a result, our model achieves an 82.18% mean average precision for class detection and a 43.4% accuracy for species detection. Furthermore, the model has the capability to identify new species through the computation of inter-species distances.

Mesoscale eddies are circular water currents found widely in the ocean and significantly impact the ocean’s circulation, water distribution, and biology. However, our comprehension of eddies’ three-dimensional (3D) structures remains constrained due to the scarcity of in-situ data. Therefore, we introduce a novel deep learning model, 3D-EddyNet, designed for reconstructing the 3D thermohaline structure of mesoscale eddies. Utilizing multi-source satellite data and Argo profiles collected from eddies in the North Pacific Ocean between 2000 and 2015, we optimized the 3D-EddyNet model by adjusting image sizes, introducing a Convolutional Block Attention Module, and incorporating eddy physical parameters. Results demonstrate remarkable accuracy, with an average root mean square error (RMSE) of 0.32 °C (0.03 psu) for temperature (salinity) within anticyclonic eddies and 0.41 °C (0.04 psu) within cyclonic eddies in the upper 1000 m. We applied 3D-EddyNet to reconstruct 3D eddy structures in the Kuroshio Extension (KE) and the Oyashio Current (OC) regions, demonstrating its capability to accurately represent the 3D thermohaline eddy structures both vertically and horizontally. The consistency in the averaged 3D eddy structures between our 3D-EddyNet and the ARMOR3D dataset in the KE and OC regions underscores the robust generalizability of our model, indicating the model’s ability to infer 3D eddy structures when Argo profiles are unavailable. The distinctive advantage offered by 3D-EddyNet enhances our ability to understand mesoscale eddy dynamics, overcoming challenges posed by the limited availability of in-situ data.

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.

Precipitation forecasting, encompassing high-resolution regional and accurate intensity forecasting, has long been a central focus in the artificial intelligence (AI) field. Generally, AI models forecast future radar precipitation sequences from historical radar precipitation sequences. However, current precipitation forecasting suffers from three issues: 1) mismatched precipitation patterns, for example, motion speed; 2) blurred precipitation field generation; and 3) inaccurate intensity forecasts. It is owing that AI models: 1) tend to simulate the average motion states and overlook individual motion (defined as the description of the motion speed, trajectory, and direction for a single precipitation sample), leading to either overestimation or underestimation of the motion speed for a specific precipitation field and 2) are inclined to filter out high-frequency components to reduce noise in the information transmission, resulting in low resolution and blurred precipitation fields. The modeling challenges impose constraints on achieving high-resolution regional and accurate intensity forecasting. Here, to the former, we propose an individual motion-driven AI (IMD-AI) method, incorporating motion matching, alignment, and refinement through constructed pattern groups. It effectively solves the mismatch in motion estimation for an individual precipitation field, ensuring global and intact regional precipitation forecasting. To the latter, we put forward a schedule sampling, patch embedding, and adversarial strategies (SPA) training mechanism, which eliminates sharp and local information loss issues during the filtering of high-frequency components. Extensive experimental results demonstrate that IMD-AI achieves accurate motion estimation of individual precipitation fields and generates high-resolution regional and accurate intensity forecasting.

Characterizing global phytoplankton pigment profiles in the ocean is crucial for comprehending phytoplankton dynamics. This study develops a self-attention-based deep learning model (Pigmentsformer) to estimate nine types of global phytoplankton pigment profiles in the top 300-m ocean. Pigmentsformer employed 15-day sequences of ocean-color data to capture the dynamic changes in phytoplankton growth. The model was trained on an extensive collection of matchups, including 33 975 samples between global high-performance liquid chromatography (HPLC) in situ measurements, ocean-color satellite measurements, and the environment fields from the reanalysis dataset. Validation of the model employs a tenfold leave-one-out cross-validation (LOOCV) approach. The coefficient of determination (${R}^{2}$) between in situ HPLC and Pigmentsformer-estimated concentrations ranges from 0.67 to 0.87, with the mean absolute error (MAE) ranging from 0.01 to$0.33~{\text {mg}\cdot }{\text {m}}^{-3}$for nine types of pigments. The backpropagation technique reveals that among all predictors, optical properties are paramount when estimating total chlorophyll-a (TChla) at the surface, with ocean current being the most influential environmental factor. However, as depth increases, the effect of environmental variables SSH and temperature exceed that of optical properties and ocean current. An examination of 20 years of model-generated phytoplankton size classes (PSCs) was conducted to explore the correlation between changes in phytoplankton communities and the El Niño-Southern Oscillation (ENSO) in the Equatorial Pacific. The location of the maximum phytoplankton layer has a positive relationship with Niño 3.4 index ($R =0.70$for micro-phytoplankton,$R =0.68$for nano-phytoplankton, and$R =0.45$for pico-phytoplankton) within the Equatorial Pacific of the Niño 3.4 region.

The acquisition of wide geophysical data of vast oceans by satellites can be impeded by clouds, which may result in gaps in the data acquired in the infrared and visual bands, such as sea surface temperature (SST) data, leading to limited data usage. To address this issue, we proposed a general and robust method, named the empirical function (EF) method, which involves expanding an ocean field with space-time-separation functions and determining the functions by minimizing the expansion’s residual on the observation data while considering the prior-knowledge constraint that an ocean field varies smoothly in space and time. To test the effectiveness of the EF method, we applied it to reconstruct the 14-year cloud-free SST data in the Gulf Stream region spanning from 24.5°N to 44°N and 82.5°W to 54.5°W. The original data consists of the 0.025°×0.025° gridded daily composite daytime SST products of the Moderate-Resolution Imaging Spectroradiometer on the Aqua sun-synchronous satellite, with an annual data-missing rate fluctuating around 78% in the region. In addition, we validated the reconstructed data against in-situ buoy measurements. The reconstructed SST data’s accuracies are -0.11 ± 0.91°C (bias ± standard deviation of error) and -0.12 ± 0.67°C in the areas without and with satellite observations, respectively, which are slightly lower and higher than the gappy satellite SST products’ accuracy of -0.14±0.77°C.

Object detection algorithms, a popular research direction in computer vision, are valued for their wide range of applications. However, its applications in marine deep-sea biology are rarely seen. Three reasons are hidden among them: first, the lack of marine deep-sea datasets; second, the difficulty of network detection units to meet taxonomic requirements (both in speed and accuracy); and third, the difficulty of judging the network as a whole. To solve these problems to achieve high accuracy and high fineness for deep-sea target detection tasks, this paper proposes a creature detection model (CDM) based on YOLO V7 tiny network incorporating the idea of a bag of features. The network proposes a complementary LOSS function to optimize the model error caused by the polarization of underwater robots (ROVs) during deep-sea operations and uses an unsupervised clustering algorithm to generate a new dictionary for underwater species classification based on the detection.

In this study, we proposed an automatic internal wave signature extraction method for SAR (Synthetic Aperture Radar) imagery based on deep convolutional neural networks (DCNNs). For model training, we collected 116 labeled ENVISAT (Environmental Satellite) ASAR (Advanced SAR) images with clear internal wave signatures in the northern South China Sea. The specially tailored deep-learning-based extraction network leverages three modifications of U-Net to improve the accuracy and robustness. As a result, the overall mean Precision, Recall, and F1-score of the model are 84.95%, 84.71%, and 84.83%, respectively. The statistical results imply that our method can significantly increase the internal wave signature extraction accuracy from SAR images, even under complicated imaging conditions.

Global navigation satellite system reflectometry (GNSS-R) Delay-Doppler map measures the sea surface roughness, which has recently been applied to retrieve sea surface wind speed. However, current studies on GNSS-R wind speed retrieval only use the spatial domain of the delay-Doppler map without considering the variations patterns in the map, which is regarded as frequency domain information of the map. In this study, we propose a joint frequency-spatial-domain wind speed retrieval network (FSNet) based on reflectivity data provided by the Cyclone Global Navigation Satellite System (CyGNSS) mission. We construct a matchup dataset between the CyGNSS satellite data and ECMWF model data from January 1, 2018, to December 31, 2019. The wind speed range is 0–25 m/s. Using the proposed FSNet, frequency and spatial features are simultaneously extracted. The frequency domain feature supplements the spatial-domain information of the mid and high-level features in the neural network. Rather than directly concatenating the frequency-domain features with the spatial-domain features, we designed a feature fusion module to fuse frequency and spatial features for wind speed retrieval adaptively. Experiments show that our FSNet wind speed retrieval has a root mean square error (RMSE) of 1.63 m/s for a wind range of 0-25 m/s. This accuracy is 25.4% better than the operational algorithm provided by the CyGNSS Level 2 wind speed product. For a higher wind range of 16-25m/s, FSNet performed even better, improving the RMSE by 31%.

During the melting season, predicting the daily sea ice concentration (SIC) of the Pan-Arctic at a subseasonal scale is strongly required for economic activities and a challenging task for current studies. We propose a deep-learning-based data-driven model to predict the 90 days SIC of the Pan-Arctic, named SICNet90. SICNet90takes the historical 60 days’ SIC and its anomaly and outputs the SIC of the next 90 days. We design a physically constrained loss function, normalized integrated ice-edge error (NIIEE), to constrain the SICNet$_{\mathrm {90{'}s}}$optimization by the spatial morphology of SIC. The satellite-observed SIC trains (1991–2011/1997–2017) and tests the model (2012/2018–2020). For each test year, a 90-day SIC prediction is made daily from May 1 to July 2. The binary accuracy (BACC) of sea ice extent (SIC$>$15%) and the mean absolute error (MAE) are evaluation metrics. Experiments show that SICNet90significantly outperforms the Climatology benchmark on 90 days prediction, with a BACC/MAE improvement/reduction of 5.41%/1.35%. The data-driven model shows a late-spring-early-summer predictability barrier (around June 20) and a prediction challenge (around July 10), consistent with SIC’s autocorrelation. The NIIEE loss optimizes the predictability barrier/challenge with a BACC increase of 4%. Using a 60 days historical SIC to predict 90 days SIC is better than a historical SIC of 30/90 days. The historical 2-m surface air temperature shows positive contributions to the prediction made from May 1 to mid-June, but negative contributions to the prediction made after mid-June. The historical sea surface temperature and 500 hp geopotential height show negative contributions.

Internal waves (IWs) are a common characteristic of oceans and serve a crucial role in transmitting energies between large-scale tides and small-scale mixing. This study developed a deep-learning-based method for extracting IW signatures on multiple satellite imagery from synthetic aperture radar (SAR) and optical sensors in sun-synchronous and geostationary orbits with varying spatial resolution. We collected 1115 satellite images, including 116 ENVISAT ASAR (Advanced SAR), 839 MODIS (MODerate-resolution Imaging Spectroradiometer), and 160 Himawari-8 AHI (Advanced Himawari Imager) images with clear IW signatures in the South China Sea (SCS), Sulu Sea, and Celebes Sea for model training. Considering the distinct IW characteristics under different imaging mechanisms, the specially tailored IW Extraction network (IWE-Net) leverages three modifications to improve the accuracy and robustness: online data augmentation, squeeze and excitation blocks, and Matthews correlation coefficient loss. The overall mean Precision, Recall, and F1-score of the IWE-Net model are 85.75%, 85.67%, and 85.71%, demonstrating the model is accurate for IW signature extraction.We also proved the transferability of our method to sea areas worldwide, long-term periods, and Sentinel satellite sensors completely independent of the model training. Globally, the number of IW images and extracted pixels show an obvious tidal-related double-peak distribution. Furthermore, we processed 15461 MODIS images in the northeastern SCS to present a holistic IW distribution map over the past 22 years. An unreported IW silent zone caused by drastic topography changes has been discovered, indicating the great potential of deep learning in information retrieval from remote sensing imagery.

Hyperspectral band selection (BS) is an important task for hyperspectral image (HSI) processing, which aims to select a discriminative and low-redundant band subset. As a significant cue for BS, structure information describes the cross-band correlation which brings the redundancy of HSI. Existing methods model structure information via manual rule-based graph construction. However, such a graph construction method fails to model complex and diverse structural relationships of HSI data. To address this problem, we propose a data-driven method, named iterative graph auto-encoder for band selection (IGAEBS). It adaptively captures structure information by a data-specific automatic construction process, rather than by a fixed empirical design. Specifically, we propose a new unsupervised pretext task to train graph convolution neural network to extract HSI features. These features are used to construct a graph to represent the structural relationships among bands. To enhance the reliability of the graph, we further design an iterative graph improvement mechanism to progressively refine the structure representation. Using the derived graph, we partition the bands into several clusters and select a representative band from each cluster. During the selection process, both intra- and inter-cluster information are considered to improve the discriminativeness of band subset. Extensive experiments are conducted on three public data sets to validate the superiority of the proposed method compared to other state-of-the-art methods.

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 develops a deep learning (DL) model to classify the sea ice and open water from synthetic aperture radar (SAR) images. We use the U-Net, a well-known fully convolutional network (FCN) for pixel-level segmentation, as the model backbone. We employ a DL-based feature extracting model, ResNet-34, as the encoder of the U-Net. To achieve high accuracy classifications, we integrate the dual-attention mechanism into the original U-Net to improve the feature representations, forming a dual-attention U-Net model (DAU-Net). The SAR images are obtained from Sentinel-1A. The dual-polarized information and the incident angle of SAR images are model inputs. We used 15 dual-polarized images acquired near the Bering Sea to train the model and employ the other three images to test the model. Experiments show that the DAU-Net could achieve pixel-level classification; the dual-attention mechanism can improve the classification accuracy. Compared with the original U-Net, DAU-Net improves the intersection over union (IoU) by 7.48.% points, 0.96.% points, and 0.83.% points on three test images. Compared with the recently published model DenseNetFCN, the three improvement IoU values of DAU-Net are 3.04.% points, 2.53.% points, and 2.26.% points, respectively.

This study developed a textural-enhanced deep learning (DL) model based on the classic U-net framework for green algae detection in Sentinel-1 synthetic aperture radar (SAR) imagery. Four special modifications are made in the framework: texture-fused input dataset, texture concatenation to effectively use the texture information, weighted loss function to settle the imbalance of algae-seawater samples, and an attention module to facilitate model focus on the discriminative features efficiently. To build the proposed model, we collected 119 Sentinel-1 SAR images acquired in the Yellow Sea and manually labeled 8441 samples, among which 4421/1896/2124 were used as the training/validation/testing dataset. Experiments show that the classification achieves the mean intersection over union (mIOU) of 86.31%, outperforming previous DL methods. Furthermore, each modification is effective, and the weighted loss function plays the most critical role. Moreover, we monitored green tide in the Yellow Sea from 2019 to 2021 using the proposed model and analyzed the relationship between green tide interannual variation and two primary environmental factors: nitrate concentration and sea surface temperature (SST). The interannual variation is characterized via three crucial indexes: bloom duration, coverage area, and nearshore damage. The detection results reveal that the bloom duration is the longest (shortest) in 2019 (2020), corresponding to the biggest (smallest) coverage area in 2019 (2020). In addition, the nearshore damage is the heaviest (lightest) in 2021 (2020). We also found that the interannual variation of green tide scales is partly related to the available nitrate concentration and SST variation in algae-distributed regions.

Satellite-based synthetic aperture radar (SAR) can provide a low-cost, frequent environment monitoring for dynamic intertidal zones. The critical problem is to realize pixel-level classification of SAR images of the intertidal zones with excellent and robust performance. Recently, deep learning, in particular deep convolutional neural networks, has provided us with promising solutions to this problem. Based on a sophisticated deep learning-based pixel-level classification model U2-Net, we propose an MB-U2-ACNet model suitable for intertidal zone land cover classification using dual-polarimetric SAR data integrated with environmental information, such as wind speed and tide level information. The MB-U2-ACNet model has a multi-branch nested U-shaped encoding-decoding structure. We extract and fuse features from multiple data sources, including satellite remote sensing and environmental information, by establishing the multi-branch structure. Furthermore, we propose an asymmetric convolution residual U-block for each encoding-decoding stage to improve the model’s feature extraction ability. Moreover, the model with attention mechanisms better distinguishes the importance of features from the channel’s perspectives and spatial dimensions. We construct a dataset with 106 Sentinel-1 SAR images from 2016 to 2020 for environment monitoring in the intertidal zone of Shanghai Nanhui. On the dataset, the proposed model reaches the overall classification accuracy of 96.40% and the mean intersection over union score of 0.8307. The experiments show the advantages of the proposed model compared with the benchmarking models due to better feature extraction and multi-source information fusion. In addition, the contributions of every added sub-structure are analyzed systematically.

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.

This study develops a deep learning (DL) model to extract the ship size from Sentinel-1 synthetic aperture radar (SAR) images, named SSENet. We employ a single shot multibox detector (SSD)-based model to generate a rotatable bounding box (RBB) for the ship. We design a deep-neural-network (DNN)-based regression model to estimate the accurate ship size. The hybrid inputs to the DNN-based model include the initial ship size and orientation angle obtained from the RBB and the abstracted features extracted from the input SAR image. We design a custom loss function named mean scaled square error (MSSE) to optimize the DNN-based model. The DNN-based model is concatenated with the SSD-based model to form the integrated SSENet. We employ a subset of the OpenSARShip, a data set dedicated to Sentinel-1 ship interpretation, to train and test SSENet. The training/testing data set includes 1500/390 ship samples. Experiments show that SSENet is capable of extracting the ship size from SAR images end to end. The mean absolute errors (MAEs) are under 0.8 pixels, and their length and width are 7.88 and 2.23 m, respectively. The hybrid input significantly improves the model performance. The MSSE reduces the MAE of length by nearly 1 m and increases the MAE of width by 0.03m compared to the mean square error (MSE) loss function. Compared with the well-performed gradient boosting regression (GBR) model, SSENet reduces the MAE of length by nearly 2 m (18.68%) and that of width by 0.06 m (2.51%). SSENet shows robustness on different training/testing sets.

This study proposes a purely data-driven model for the weekly prediction of daily sea ice concentration (SIC) of the pan-Arctic (90 N, 45 N, 180 E, 180 W) during the melting season. The model, SICNet, adopts an encoder–decoder framework with fully convolutional networks (FCNs) and can predict the SIC (covering$320\times224$grids, each with a resolution of 25 km) one-week lead with high accuracy. We design a temporal–spatial attention module (TSAM) to help SICNet capture spatiotemporal dependencies from SIC sequences. The satellite-derived SIC data of 33 years (1988–2020) from the National Snow and Ice Data Center (NSIDC) are employed to train and test the model, 1988–2015 for training, and 2016–2020 for testing. SICNet achieves the mean absolute error (MAE) of 2.67%, the mean absolute percentage error (MAPE) of 8.67%, and the Nash–Sutcliffe efficiency (NSE) of 0.9784 in weekly predicting of SIC during the melting season. SICNet achieves better performance than existing deep-learning-based models. The TSAM reduced the MAE from 2.73% to 2.67%. We evaluate the model’s performance by recursively predicting, from seven- to 28-day leads. We employ the binary accuracy (BACC) metric to measure the accuracy of the predicted sea ice extent (SIE) and compare SICNet with the anomaly persistence (Persist). SICNet shows better performance than Persist with an average BACC on the 28th day of 2016–2019 over 90% (90.17%). For the 28-day lead predictions of three extreme minimum SIE in September 2007, 2012, and 2020, SICNet outperforms Persist with an average improvement of 1.84% in BACC and$0.16 milkm^{2}$in the SIE error.

In this study, a set of deep convolutional neural networks (CNNs) was designed for estimating the intensity of tropical cyclones (TCs) over the Northwest Pacific Ocean from the brightness temperature data observed by the Advanced Himawari Imager onboard the Himawari-8 geostationary satellite. We used 97 TC cases from 2015 to 2018 to train the CNN models. Several models with different inputs and parameters are designed. A comparative study showed that the selection of different infrared (IR) channels has a significant impact on the performance of the TC intensity estimate from the CNN models. Compared with the ground truth Best Track data of the maximum sustained wind speed, with a combination of four channels of data as input, the best multicategory CNN classification model has generated a fairly good accuracy (84.8%) and low root mean square error (RMSE, 5.24 m/s) and mean bias (−2.15 m/s) in TC intensity estimation. Adding attention layers after the input layer in the CNN helps to improve the model accuracy. The model is quite stable even with the influence of image noise. To reduce the side-effect of the very unbalanced distribution of TC category samples, we introduced a focal_loss function into the CNN model. After we transformed the multiclassification problem into a binary classification problem, the accuracy increased to 88.9%, and the RMSE and the mean bias are significantly reduced to 4.62 and −0.76 m/s, respectively. The results show that our CNN models are robust in estimating TC intensity from geostationary satellite images.

Synthetic aperture radar (SAR) can monitor the sea surface imprints of tropical cyclones (TCs) with high spatial resolution, day and night. Automatically locating TC center positions in SAR images is a challenging task. This article developed a two-stage, fully automatic TC-center estimation algorithm. First, the sea surface wind directions (SSWDs) at SSWD points are retrieved by the improved local gradient (ILG) method. We incrementally deflected the SSWD outward at a 0.5° angle from −50° to 10° (the negative angles represent clockwise deflection). The heat maps are generated for each of the 121 angles, and the values at each heat map are the cumulative numbers of the lines perpendicular to the compensated SSWDs. The site corresponding to the maximum cumulative number in all 121 heat maps is the coarsely estimated center position. This center search is the culmination if it falls outside the SAR image. Otherwise, the second stage is triggered, and the sub-SAR image (150 km$\times150$km) centered at the coarsely estimated center position is extracted. Then, the first-stage procedure is repeated with the sub-SAR image to precisely estimate the center position. Optionally, the precisely estimated center position can be further adjusted by considering that normalized radar cross section (NRCS) is normally minimal at the TC center. We applied the algorithm to 87 SAR images. Five of these images do not contain TC centers. The results are in good agreement with the visually located TC center positions and those in the best track (BT) datasets.

We propose a methodology to estimate the tropical cyclone (TC) center location associated with the additional TC parameters of the radius of maximum wind (RMW) and intensity, purely from ocean winds observed by the Soil Moisture Active Passive (SMAP) radiometer. This method assumes a symmetric vortex. To demonstrate the method, we analyze 28 SMAP wind fields collected during 11 TCs. Verification of the estimated hurricane centers and wind distributions along the radius is compared with measurements provided by the airborne stepped-frequency microwave radiometer (SFMR) collected during its flying through hurricane cores and to sea surface wind fields derived from spaceborne synthetic aperture radar data. The significance of this simple method is that TC centers, intensities, and RMW can be estimated purely from SMAP wind products, despite SMAP’s low spatial resolution. Comparing these 28 model results to the aircraft measurement (f-deck) and the best track (BT) data, we show that, for the latitudes and longitudes of the TC centers, the standard deviations are 0.28° and 0.29°, respectively, both with a correlation of 1.00. For the detected intensity, the standard deviation is 8.11 m/s, and the correlation is 0.84. We note that the TC intensity detected by this method can be even stronger than the maximum winds observed by the relatively low-resolution SMAP observations.

Sea fog significantly threatens the safety of maritime activities. This paper develops a sea fog dataset (SFDD) and a dual branch sea fog detection network (DB-SFNet). We investigate all the observed sea fog events in the Yellow Sea and the Bohai Sea (118.1°E-128.1°E, 29.5°N-43.8°N) from 2010 to 2020, and collect the sea fog images for each event from the Geostationary Ocean Color Imager (GOCI) to comprise the dataset SFDD. The location of the sea fog in each image in SFDD is accurately marked. The proposed dataset is characterized by a long-time span, large number of samples, and accurate labeling, that can substantially improve the robustness of various sea fog detection models. Furthermore, this paper proposes a dual branch sea fog detection network to achieve accurate and holistic sea fog detection. The poporsed DB-SFNet is composed of a knowledge extraction module and a dual branch optional encoding decoding module. The two modules jointly extracts discriminative features from both visual and statistical domain. Experiments show promising sea fog detection results with an F1-score of 0.77 and a critical success index of 0.63. Compared with existing advanced deep learning networks, DB-SFNet is superior in detection performance and stability, particularly in the mixed cloud and fog areas.

Change detection for remote sensing images involves detecting regional surface changes of interest between two images taken of the same geographical area but at different times. In image processing, the spatial domain uses grayscale values to describe an image. The frequency is directly related to the spatial change rate, so the frequency domain can be intuitively associated with patterns of intensity variations in the image. These two domains provide different perspectives for image interpretation. Most existing deep-learning-based methods formulate change detection as a pixel-wise binary classification problem and utilize various strategies to extract information in the spatial domain. However, they rarely pay attention to the rich information in the frequency domain. To address this problem, we propose an end-to-end joint frequency-spatial domain network (JFSDNet) to implement remote sensing optical image change detection. Specifically, we introduce frequency information into the change detection to supplement the loss of image details caused by down-sampling. In addition, we employ a frequency selection module to adaptively discriminate and choose frequency clues by reducing the complexity of the frequency features. The JFSDNet is applied to two publicly available datasets: the CDD dataset and the LEVIR-CD dataset. Compared with other methods, both visual interpretation and quantitative assessment confirmed that our proposed method achieved a favorable performance.

Sea surface height anomaly (SSHA) is vitally important for climate and marine ecosystems. This article develops a multilayer fusion recurrent neural network (MLFrnn) to achieve an accurate and holistic prediction of the SSHA field, given only as a series of past SSHA observations. The proposed approach learns long-term dependencies within the SSHA time series and spatial correlations among neighboring and remote regions. A new multilayer fusion cell as the building block of the MLFrnn model was designed, which fully fused spatial and temporal features. The daily average satellite altimeter SSHA data in the South China Sea from January 1, 2001, to May 13, 2019, were used to train and test the model. We performed a 21-day ahead SSHA prediction and our MLFrnn model has very high accuracy, with a root mean square error (RMSE) of 0.027 m. Compared with existing deep learning networks, the proposed model was superior both in prediction performance and stability, especially on the wide-scale and long-term predictions.

Wind-induced oriented textures (WIOTs) are commonly used to retrieve sea surface wind directions from synthetic aperture radar (SAR) images. In this study, we found that WIOTs are also related to sea surface wind speeds (SSWSs). The entropy values in the gray-level cooccurrence matrices (GLCMs) for SAR images containing WIOTs will become steady with increasing distance between pairs of pixels. Furthermore, these steady values of entropy (SVEs) show a clear linear relationship with SSWSs. As a result, an SSWS retrieval model was developed based on this relationship. We used 2222/2223 Sentinel-1 SAR images (wind speed ranges from 5 to 20 m/s) to fit/validate the algorithm. The retrieved SSWSs were compared with the European Centre for Medium-Range Weather Forecast (ECMWF) SSWSs, Cross-Calibrated Multi-Platform (CCMP) SSWSs, and Tropical Atmosphere/Ocean (TAO) buoy measurements, and the root-mean-square differences (RMSDs) were 1.78, 1.70, and 1.78 m/s, respectively. The new model was also tested for SAR images acquired under hurricane conditions. The wind comparisons against stepped-frequency microwave radiometer (SFMR) measurements show an RMSD of 1.28 m/s. Our model’s performance was also tested with the images at different spatial scales in the validation data set. Since the model is based on inherent image patterns, it still works well for SAR images without precise calibration.

Sea ice type's classification plays an essential role in Arctic marine navigation. Synthetic Aperture Radar (SAR) is independent of weather conditions and is widely used in sea ice classification. U-Net is a well-performed deep learning (DL) framework in image classification. This study constructs a U-Net-based “end-to-end” model to classify multi-year ice (MYI), first-year ice (FYI), and open water. We label the SAR images by ice chart provided by the U.S. National Ice Center (USNIC). The labeled images are divided into chips to be fed into the U-Net model for training and testing. Experiments show that the precision and the recall of the testing set are 89.55% and 89.46%. The proposed model can classify sea ice types in an “end-to-end” way with high accuracy.

Recently, deep learning has already shown its availability in synthetic aperture radar (SAR) image classification. To improve the deep learning model's performance on multisource remote sensing information fusion, we propose a multi-branch deep convolutional neural network model specially tailored from the U2-Net framework and with asymmetric convolutions. We name it the MB-U2-ACNet model. Based on experiments on a constructed dataset dedicated for Shanghai Nanhui Dongtan intertidal zone environment monitoring, it is verified that the proposed MB-U2-ACNet model has better performance than the existing representative deep and traditional methods.

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.

This study proposes a fully-convolutional-networks-based (FCN-based) model to predict the daily Arctic SIC in the melt season. The FCN-based model adopts an encoder-decoder framework with FCN layers as the basic unit. We use the SIC series of the last seven days to predict the SIC of the coming day. The Arctic SIC series of 31 years (1988–2018) from the NSIDC are employed to train and evaluate the model. Experiments show that the FCN-based model is capable of accurately predicting the daily Arctic SIC of all grids (320× 224) in an end-to-end way, with an MAE under 1%. The FCN-based model shows apparent advantages over the newly published convolution-neural-network-based (CNN-based) model in prediction accuracy, efficiency, and resource occupation.

The effects of sea surface temperature (SST) on ocean radar backscattering are investigated under both the neutral and nonneutral atmospheric conditions for the applications of wind retrieval. The impact factors are parameterized as functions of SST. The SST effects on the variations in ocean scattering and wind retrieval are evaluated using an analytic model which combines the KHCC03 spectrum and the second-order small slope approximation (SSA-II) model. Under the neutral condition, we present the following new insights at three commonly used bands: 1) the seawater permittivity accounts for a dominant effect of SST at the L-band. The SST effects induce a wind underestimation of 0.3 m/s over cold seawater and a wind overestimation of 0.24 m/s over warm seawater at the L-band and a wind speed of 8 m/s. The seawater viscosity plays a significant role in the SST effects on ocean scattering at the C- and Ku-bands, while its variation induced by SST has insignificant effects on L-band scattering; 2) for the C-band, the SST-induced wind retrieval error can be neglected at a medium wind speed due to the neutralization of the effects of various factors on surface roughness. Yet, the SST effects are not negligible at low and high wind speeds; and 3) both the dielectric and dynamic factors play significant roles in the SST effects on ocean scattering at the Ku-band. Under the nonneutral condition, the simulation results show that the air–sea interaction governs the SST effects on ocean scattering and wind velocity variations. Other than the air–sea interaction, the wind retrieval errors induced by other SST-related factors are negligible at the L- and C-bands.

Internal waves (IWs), observed in the world oceans, have significant impacts on ocean engineering and environments. In this study, we collected satellite images from Moderate-Resolution Imaging Spectroradiometer and Visible Infrared Imaging Radiometer Suite sensors in the Sulu-Celebes Sea from 2016 to 2019 to understand the IW generation and propagation. Satellite observations show a coherent IW phase difference in both seas, indicating that the IWs are alternatively generated when the tidal currents oscillate back and forth in the Sulu Archipelago, which separates two seas. A new generation site is found for occasionally observed long IWs in the eastern Sulu Sea. To understand the IW propagation characteristics, we developed a machine-learning-based forecast model. We trained the model with both IW wave crest curves extracted from satellite images and published climatological ocean temperature–salinity profiles. Since many satellite images contain IW packets generated at two or three tidal cycles, we can validate the model performance by matching the model prediction after one or two tidal cycles with the second or third wave crests in satellite images. Three factors are adopted to evaluate the forecast results: the root-mean-square error (RMSE), the Fréchet distance (FD), and the correlation coefficient (CC). The forecast model has an average error with an RMSE of 12.92 km, an FD of 18.73 km, and a CC of 0.98. Analysis shows that a smaller time step is preferred in regions where the water depth changes significantly. Comparison with the Korteweg–de Vries equation solutions shows that the developed forecast model is more robust when errors introduced to the model inputs.

The effects of sea surface temperature (SST) on ocean radar backscattering and wind retrieval are investigated using the second-order small slope approximation (SSA-II) model. Impact factors are parameterized and built into an SST-enhanced KHCC03 spectrum. By employing the Monin-Obukhov similarity theory (MOST), the air-sea interaction is considered in the analyses. Under the neutral condition, the SST effects on wind retrieval cannot be neglected at L-band and the seawater permittivity and air density are the dominant factors. For C-band, the SST-induced wind retrieval error can be neglected at a medium wind speed due to the neutralization of the effects of various factors on surface roughness. Yet, the SST effects are not negligible at low and high wind speeds. At Ku-band, the SST-related factors besides the air density and seawater viscosity can also significantly affect the ocean backscattering.

Mesoscale eddies are circular flowing currents that can retain and transport salt, heat, and nutrients all around the ocean. Mesoscale eddies can be detected on remote sensing images, i.e., sea surface height (SSH) images, sea surface temperature (SST) images, chlorophyll concentration images, etc. Most existing automatic eddy detection algorithms are developed based on one kind of remote sensing data. There is a lack of an automatic eddy detection algorithm that can make full use of multi-source remote sensing data to ensure the accuracy and efficiency of eddy detection. The paper proposes a multi-modal U-Net model, a deep neural network-based framework for eddy detection from multi-source remote sensing images. Compared with the previous eddy detection methods, the newly proposed method improves the accuracy and efficiency of eddy detection by using fusion data of SSH and SST.

Researchers have already verified that the deep learning (DL) technology can realize accurate and robust mapping of tropical cyclone-induced coastal inundation in synthetic aperture radar imagery. In order to liberate the DL-based inundation mapping from human supervision, we propose to use the clustering of deep convolutional autoencoder-generated features. The mapping results of Lekima 2019-induced inundation demonstrate the advantages and availability of the proposed method.

Accurate and robust classification methods of sea ice and open water are significant for many applications. Synthetic Aperture Radar (SAR) imaging capability is independent of weather conditions and is widely used in sea ice classification. U-Net, a deep learning framework, has achieved great success in the field of biomedical image classification. In this study, we construct a U-Net-based “end-to-end” model to classify the sea ice and open water pixels in SAR imagery. Five SAR images acquired in the Gulf of Alaska near Bering Strait are used in this case study. We manually label the SAR images as ice and water. The labeled images from the first four SAR image are divided into chips to be fed into the U-Net model for training. The fifth SAR image is employed as the testing data. Experiments show that the precision and the recall of the testing image is 91.64% and 91.70%, respectively. Most of the sea ice, including small chunks and sinuous ice edges, can be successfully classified.

In this study, a deep convolutional neural network (CNN) was developed to forecast the movement direction of tropical cyclones (or typhoons) over the Northwestern Pacific basin from Himawari-8 (H-8) satellite images. 2250 infrared images which captured 97 typhoon cases between 2015 and 2018 were used to train the CNN model. By using images from Channels 13 and 15 as input into the CNN model, the mean error of the typhoon movement angle reaches up to 27.8°, which shows the great potential of deep learning in tropical cyclone track prediction.

In this study, we proposed an automatic internal wave (IW) signature extraction method based on the deep convolutional neural networks (DCNN). Our objective is to provide a rapid and simple to use method that can tackle the IW signature extraction in images from different satellite sensors without re-training or manual interference. We proved the generalization ability of our method across multiple optical satellite sensors. The statistical results show this DCNN-based method has appreciable transferability and is promising for efficient extraction of internal wave signature in different satellite images with varying spatial resolution even under complex imaging conditions.

The sea surface wind at low and moderate speed (20 m/s). In this study, we explore to use a novel deep convolutional neural networks (DCNN) architecture, U-Net, to retrieve the sea surface wind at high speed. The Sentinel-1B SAR cross-polarized (HV and VH) Normalized Radar Cross Section (NRCS), SAR incidence angle and the ancillary Global Forecast System (GFS) model wind directions are used as the U-Net input data. The GFS model wind speed is the ground-truth data. The results suggest that the U-Net wind retrieval model has a capability to retrieve the high wind speed from SAR data with the physical model-based data. The accuracy of physical model data directly affects the U-Net model results. Meanwhile, U-Net model has a better continuity in the joint area of two strips.

In this study, a deep convolutional neural network was designed to estimate the intensity of tropical cyclones over the Northwestern Pacific Ocean from high-frequency Himawari-8 satellite images. Our model achieved good results by using the brightness temperature derived from one single infrared band data. The accuracy of the top (top-1) and the second best (top-2) tropical cyclone intensity classification reaches 81.4% and 93.3%, respectively.

In the past few years, artificial intelligent (AI) technology has been widely used in many research fields for big data information mining and shown great potential applications in computer vision, natural language processing, bioinformatics, among others. In the area of remote sensing oceanography, we categorize its applications in four major categories: image classification, data fusion, algorithm development and oceanic phenomenon forecast. In this paper, we present two examples to demonstrate such applications. In the first example we applied a well-studied AI framework, U-Net, to a NASA JPL’s UAVSAR Synthetic Aperture Radar (SAR) image to classify coastal zone types in the Gulf Coast of USA. In the second example, we trained the AI framework, LSTM model, using the time series of blended microwave Sea Surface Temperature (SST) data, and made the equatorial SST pattern forecast. Validation studies in both cases showed the robustness of AI-based technology for oceanography research.

Synthetic aperture radar (SAR) has emerged as a new tool for tropical cyclone (TC) monitoring by providing information on the location of TC centers. However, SAR does not usually cover the entire TC domain due to its limited swath width. In this paper, we develop a procedure to identify the location of the center of a TC when an SAR image only covers the rain band portion of the TC but not the eye. The algorithm is based on both an image processing procedure and the available knowledge of the inherent rain-band structure of a TC. The three-step algorithm includes: 1) applying a Canny edge detector to find the curves associated with rain bands; 2) defining two filter criteria to select the spiral curves that resemble the estimation based on a TC rain-band model; 3) searching for the optimal matching solution using the particle swarm optimization algorithm. Numerical experiments with images without TC eye information show that the proposed method can effectively locate the centers of TCs. We compare the experimental results with the best track data to indicate the accuracy. Then, we compare the inflow angle model and the logarithmic spiral model and find that the inflow angle model is more accurate for TC center identification.

In this paper, we demonstrate that the spaceborne dual-platform TerraSAR-X (TSX) and TanDEM-X (TDX) pursuit monostatic mode full polarimetric (full-pol) synthetic aperture radar (SAR) data with a time lag can be used to monitor maritime traffic. For single polarization (single-pol) SAR data, the performance of vessel velocity estimation is mainly determined by 2-D cross correlation of SAR intensity data. As the sea clutter is changing dynamically during the TSX/TDX data acquisition, the correlation between two dual-platform images decreases significantly. We may get unstable or incorrect estimations of vessel velocity, especially under a higher wind condition. For solving this problem, we propose an object-oriented polarimetric likelihood ratio test (PolLRT) method based on the complex Wishart distribution. The proposed method makes PolLRT statistics of the detected target pixels for eliminating the effect of varied sea clutter. Two pairs of full-pol SAR data sets covering the Strait of Gibraltar acquired by dual-platform TSX/TDX in pursuit monostatic mode with a time lag of approximately 10 s are selected for the experiments. The experimental results demonstrate that the proposed PolLRT method has a better performance than that of the classical normalized cross correlation (NCC) method with VV polarization SAR data and the mutual information (MI) method with full-pol SAR data. Specifically, under the lower wind condition, the correct estimation rate of the NCC, the MI, and the proposed PolLRT methods are 85.7%, 57.1%, and 100%, respectively; under the relatively higher wind condition, the correct estimation rate of the above three methods are 48.8%, 23.2%, and 90.1%, respectively.

In this paper, we proposed a complete polarimetric covariance difference matrix [CP]-based algorithm for ship detection in polarimetric synthetic aperture radar (PolSAR) imagery. To calculate [C P], we first developed a scheme to reflect the polarimetric scattering differences between ship pixel (SP) and its neighboring pixels (ISPs) and, then, dividedly accumulated the amplitude and phase differences between SP and ISPs. Compared to the polarimetric covariance difference matrix [P] developed in our earlier work, [C P] effectively overcomes the drawback of the lack of the phase information in [P]. To demonstrate the effectiveness of the proposed algorithm, we applied the [CP]-based ship detection algorithm to four PolSAR data sets, including one UAVSAR L-band data set with 21 ships, two AIRSAR L-band data sets with 11 and 22 ships, respectively, and one Radarsat-2 C-band data set with 8 ships. Experimental results show that: (1) the proposed algorithm can effectively detect ships with high target-to-clutter ratio (TCR) values and (2) [C P] has a better performance than the traditional polarimetric covariance matrix [C] and [P] on ship detection. To be more specific, the average TCR value of the proposed algorithm (23.86 dB) is 6.07 and 7.47 dB higher than PNFC(i.e., the geometrical perturbation-polarimetric notch filter) and RSC(i.e., the reflection symmetry method), respectively.

This article presents a novel technique for automatically locating tropical cyclone (TC) centers based on top cloud motions in consecutive geostationary satellite images. The high imaging rate and spatial resolution images of the Gaofen-4 geostationary satellite enable us to derive pixel-wise top cloud motion data of TCs, and from the data, TC spiral centers can be accurately determined based on an entirely different principle from those based on static image features. First, a physical motion field decomposition is proposed to eliminate scene shift and TC migration in the motion data without requiring any auxiliary geolocation data. This decomposition does not generate the artifacts that appear in the results of the previously published motion field decomposition. Then, an algorithm of a motion direction-based index embedded in a pyramid searching structure is fully designed to automatically and effectively locate the TC centers. The test shows that the TC concentric motions are more clearly revealed after the proposed motion field decomposition and the located centers are in good agreement with the cloud pattern centers in a visual sense and also with the best track data sets of four meteorological agencies.

The brightness temperature data acquired during 2012-2015 from the scanning microwave radiometer (SMR), onboard the first Chinese ocean dynamic environment satellite- Haiyang-2A, were matched up with the WindSat Polarimetric Radiometer (WindSat) 0.25° × 0.25° gridded daily sea surface temperature (SST) data. Then, the artificial neural network (ANN) ensemble (ANNE) method implementing the Crogging technique was used to build the SMR SST retrieval algorithm. Different from a regular ANN, an ANNE combines the outputs of its ANN members to generate an algorithm. The developed ANNE algorithm for SMR SST was validated based on the SMR/WindSat data pairs that were not used in the tuning of the algorithm. The SST comparison shows the root mean square (rms) of 1.16 °C for the ANNE algorithm. We further validate the SMR SST products using the in situ measurements from the National Oceanic and Atmospheric Administration iQuam System. The rms of the ANNE algorithm in comparison with the global iQuam SSTs is 1.46 °C. All validations showed that ANNEs were more accurate than the other statistically based SST retrieval algorithms for SMR, and generally had much smaller uncertainties than regular ANNs.

In this study, polarimetric synthetic aperture radar (PoISAR)-based classification algorithms are considered to investigate the role played by polarimetric information in the classification process of coastal areas that call for heterogeneous scattering properties. Hence, a multi-frequency PolSAR dataset collected over the study area of the Yellow River delta (China) is exploited to point out benefits and limitations that characterize well-known unsupervised classification schemes. Experimental results show the potential and the drawbacks of the exploitation of multi-frequency and multi-polarization SAR measurements for challenging coastal area classification.

In this study, we exhibit the temporal variability of Sea State Bias (SSB) correction in TOPEX (sides A and B), Jason-1, Jason-2 and Jason-3 missions over 1993-2017 time span. Although the trend in long-term global mean 2D nonparametric SSB correction is not significant from zero (- 0.03±0.03 mm/yr, e.g., accounts for 1% of current global mean sea level rate) during 1993-2016, it contributes - 1.27±0.21 mm/yr and -0.26±0.13 mm/yr in TOPEX-A and Jason-2 missions, respectively. Also, the SSB trend in TOPEX-A may be partly related to the recently reported sea level trend drift during 1993-1998. On zonal average, SSB correction causes about 1% uncertainty in mean sea level trend. In regions with high significant wave height (SWH), the uncertainties grow to 2% and 4% at near 50°N and 60°S, respectively.

The marine atmospheric boundary layer (MABL) roll plays an important role in the turbulent exchange of momentum, sensible heat, and moisture throughout the boundary layer of tropical cyclones. Hence, rolls are believed to be closely related to storm development and intensification. In this study, based on the RADARSAT-2 dataset including various tropical cyclone (TC) intensities, the roll characteristics are retrieved from synthetic aperture radar (SAR) images via fast Fourier transform (FFT). We investigate the roll wavelengths at variance of TC intensities and found that the roll wavelengths are related to the distance with respect to TC center and TC intensities. This study is promising to bring roll-induced effects into hurricane forecasting model.

A gray-level co-occurrence matrix (GLCM)-based method was developed for better texture orientation estimation in remote sensing imagery. A GLCM is essentially the joint probability distribution of gray levels at the position pairs satisfying a specific relative position within an image. We first found that when the relative position is aligned with texture orientation, larger elements of the corresponding GLCM are concentrated diagonally. Then, we developed a new texture orientation estimation method. The method uses the GLCMs of relative positions equally spaced in orientation and distance, and three schemes of these GLCMs are calculated. A GLCM-derived parameter is then defined to quantitatively measure the degree of diagonal concentration of the GLCM elements, and its integral over the variable of relative distance is selected as an indicator to find the dominant texture orientation(s). For testing, we applied the method to 44 selected images containing one or multiple aligned textures. The results show that the method is in good agreement with visual inspections from 45 randomly selected people, and is insensitive to large typical noises and illumination change. In addition, using (any) one GLCM calculation scheme over the others does not significantly affect the results. Finally, the method was applied to sea surface wind direction (SSWD) retrieval from 89 synthetic aperture radar images. In the application test, the developed method achieves better SSWD retrieval accuracy than do the commonly used Fourier transform- and gradient-based methods by 8.13° and 16.09° against the European Centre for Medium-Range Weather Forecast ERA-Interim reanalysis data and 10.21° and 17.31° against the cross-calibrated multiplatform data.

This paper is dedicated to the “40 Years of Ocean Remote Sensing - a Session to Honor W. Alpers on the Occasion of his 80th Birthday”. In this paper, a list of Alpers' research contributions in understating the atmospheric gravity waves and oceanic internal waves are highlighted along with up-to-date research finding in these fields. On the atmospheric side, we presented the SAR observation of atmospheric gravity waves (AGW) generated at the lee side of a mountain ridge and the upstream propagating waves. WRF simulation is carried out to understand the generation, propagation and decay of the AGW. On the oceanic side, internal wave (IW) observations in the South China Sea are presented. We present model simulation to understand the mechanisms that cause the IW reflection.

The spatiotemporal fluctuation of flying squid resources from July to November in 2009 to 2014 and its relationship with ocean environmental variables are analyzed with the GAM model in this study. The most suitable SST, SLA and Chla are 16-18°C, 0.5-4 cm and 0.25-0.3 mg/m3, respectively. The model results showed that the Month and the Latitude have greatest effect on the CPUE. The influence of environmental factor followed by G0-100, G200-300, SST, T300, T100, SSTG, Chla, SLA and EKE.

The use of co-polarization (HH or VV) synthetic aperture radar (SAR) normalized radar cross section (NRCS) measurements to estimate winds speed has matured to an operational product. However, these wind speed retrievals have only been well validated for wind speeds less than 20 m/s and require some a priori estimated of the wind direction. At higher wind speeds, the radar cross section for co-polarization imagery becomes less sensitive to increases in wind speed. The use of cross-polarization (HV or VH) NRCS measurements is difficult because of the low NRCS values for wind speeds less than 5-10 m/s. However, as wind speeds become higher, the radar cross section becomes significantly larger than the system noise. Cross-polarization data from Radarsat-2 SAR imagery has shown promise for estimating wind speeds above 10 m/s and the retrievals can be independent of knowledge of the wind direction. In this paper, we apply a published cross-polarization wind speed algorithm to Sentinel-1 imagery. We find improved wind speed retrievals over co-polarization retrievals in very high winds especially in hurricanes. However, the noise floor in the imagery is a significant problem, especially for low wind conditions. In addition, the specification of the noise floor at beam boundaries in multi-beam imagery causes wind speed image discontinuities.

Early in the development of any field of inquiry, new and interesting phenomena are uncovered and appreciated. These may be of general scientific interest, but is it often many years before these initial observations mature to the point that they can be employed for practical applications. Dr. Werner Alpers is one of those unique individuals whose long productive career spans the evolution from research to operational applications. On the occasion of his 80th birthday, we discuss the contributions of Dr. Alpers in the context of the recent implementation of operational products based on his work and the work of others. These applications are particularly important in the field of remote sensing of the ocean from space by synthetic aperture radar (SAR).

The Category-4 typhoon Malakas moved along the continental shelf margin of the East China Sea from September 16 to 18, 2016. The typhoon-induced 2°C-3°C sea surface cooling and 0.8 mg/m3Chlorophyll-a concentration change were observed by the geostationary satellite sensors carried by the Japanese Himawari 8 Advanced Himawari Imager (AHI) and the South Korean Geostationary Ocean Color Imager (GOCI). Numerical simulation was employed to analyze this typhoon's impact on the ocean environment to the northeast of Taiwan. The model result suggests that Typhoon Malakas generated an upper layer cyclonic eddy in the study area. This surface eddy was connected to a 100-m deep pre-existing eddy to form a stronger one that triggered the upwelling observed by satellites.

In recent years, both passive and active (i.e., Synthetic Aperture Radar or SAR) satellite remote sensing has proven to be valuable tools for mapping land cover. Most of the classification algorithms are based on image-intensity and they do not perform well in different coastal zone types, because these terrains have similar optical or radar backscattering signals. In this paper we propose a weighted joint sparse on the three-channel to mine the polarimetric features and texture information. The proposed method can update weights automatically according to the difference of the three channels' contribution and the least residual error. The similarity and distinctiveness of three channels are used to deal with the complex object classification. Hybrid sparse coefficients are input to the support vector machine for fully polarimetric image classification. The proposed algorithm performed well in distinguishing some coastal land-use types. A comparison study is also conducted to show that proposed algorithm outperforms two commonly classification methods.

Satellite observations reveal that an ocean cooling event happened along the Yellow Sea coast of China intermittently in spring 2008, which lasted for days. During this period, the sea surface temperature (SST) dropped 3 °C-4 °C and the chlorophyll A (Chl-a) content increased by 0.5-1 mg/m3, as determined from satellite-derived products. The cold water also suppressed the sea surface capillary waves and made the ocean surface smooth, a distinct feature shown as dark patches observed in the synthetic aperture radar image acquired during this period of time. The surface wind direction varied between alongshore and offshore. We implemented an interactively coupled ocean (regional ocean modeling system) and atmosphere (Weather Research and Forecasting model) model to capture the dynamical processes of this seemingly wind-driven cooling event. When the wind changed direction such that the alongshore component blew with the land on its left side, stronger upwelling occurred; and when the wind blew offshore with no alongshore component, the upwelling still occurred in this area, but with less strength. Two simulations with idealized alongshore and offshore winds show that the upwelling can be set up within several hours. The alongshore wind is more effective than the offshore wind in transporting upper level water offshore and triggering upwelling and causing SST cooling areas that are relatively large in size, although the maximum SST cooling they cause is on the same order of magnitude.

Spaceborne microwave synthetic aperture radar (SAR), with its high spatial resolution, large area coverage, day/night imaging capability, and penetrating cloud capability, has been used as an important tool for tropical cyclone monitoring. The accuracy of locating tropical cyclone centers has a large impact on the accuracy of tropical cyclone track prediction. Usually, the center of a tropical cyclone can be accurately located if the tropical cyclone eye is fully covered by a SAR image. In some cases, due to the limited coverage of the SAR, only a part of a tropical cyclone can be imaged without the eye. From a SAR image processing point of view, these facts make the automatic center location of tropical cyclones a challenging work. This paper addresses the problem by proposing a semiautomatic center location method based on salient region detection and pattern matching. A salient region detection algorithm is proposed, in which the salient region map contains mainly the rain bands of a tropical cyclone in a SAR image. The pattern matching problem is transformed into an optimization problem solved by using the particle swarm optimization algorithm to search the best estimated center of a tropical cyclone. To estimate the accuracy of the located center, we compare the results with the NOAA National Hurricane Center's best track data. Experiments demonstrate that the proposed method achieves good accuracy for locating the centers of tropical cyclones from SAR images that do not contain a distinguishable eye signature.

Estuary plumes on the western Louisiana continental shelf in the Gulf of Mexico, outside of constricted channels of the Sabine and Calcasieu Lakes were observed on two RADARSAT-1 synthetic aperture radar (SAR) images. The in situ data showed a change in salinity of ~2-3 PSUs across the front on the northwest edge of the plume, corresponding to locations of vertical salinity variation of 5.5-7.5 PSU. Velocity magnitude outside of the inlet reached more than 2.3 ms-1seaward and a brackish lens in the upper 4.5 m of the water column, the spread of which contributed to the development of a plume. The volume of brackish water within the plume was estimated an order of magnitude larger than the volume output from the Calcasieu River within an entire ebb tidal period. This implies that the plume was formed during multiple tidal cycles and it sustained during the flood tides because only weak diurnal tides exist. Both satellite and in situ observations showed an interesting western intensification of the plume. An analysis to the momentum/vorticity equations suggests that the western intensification is due to the asymmetry of the vorticity-divergence relationship under the Coriolis force. This can also be enhanced by an ambient coastal current that strengthens the convergence at the front. Estuary plumes are shown as bright and dark features in two SAR images. Results from a radar model simulation and in situ water temperature data showed that the SAR observed plume features were dominated by the water temperature differences.

Two satellite synthetic aperture radar (SAR) observations of the fetch in the Bohai Sea of China are presented. The sea surface winds derived from SAR data indicated a high wind of 15-16 m/s that occurred in the fetch zone. The winds are shown to have immediate direct mechanical forcing impacts on the significant wave heights (Hs) of ocean surface gravity waves. Buoy measurements and numerical wave modeling results show that the Hs increased to a maximum of 3.5 m in the semienclosed sea, 3 h after the passage of the fetch winds, and the high Hs in the sea was sustained for a total of 6 h. The Weather Research and Forecasting (WRF) model implemented in our modeling simulation captured the wind field responsible for the evolution of the fetch event. The model-simulated surface horizontal winds agree with the SAR-derived winds. In addition, the vertical wind distribution reveals that the fetch wind field reached the 800 hPa level, and the event lasted less than one day. This study demonstrates the synergy of using SAR imagery and the WRF model as effective tools to investigate the lateral and vertical structures of coastal wind.

Sediments on exposed intertidal flats are very dynamic and perform vital ecosystem functions. This paper proposes a new classification scheme for mud and sand flats on intertidal flats using fully polarimetric synthetic aperture radar (SAR) data. Freeman-Durden (FD) and Cloude-Pottier (CP) polarimetric decomposition components as well as double bounce eigenvalue relative difference (DERD) are introduced into the feature sets instead of the original intensity polarimetric channels. Classification is carried out using the random forest (RF) theory, and the results are evaluated using confusion matrices, kappa coefficients, and RF variable importance indices. Three study sites with different environmental conditions are chosen to demonstrate the effectiveness of the proposed classification chain. To further assess the performance of the proposed feature set, we set different feature combinations and process with the same processing chain. Results show that the DERD parameter can detail the sediment mappings on exposed intertidal flats and is a useful SAR feature to distinguish mud and sand flats efficiently. The combined FD and CP components have the ability to describe the polarimetric characteristics of sediments more correctly than the commonly used original intensity channels. Meanwhile, the RF theory shows great potential in distinguishing sediments in intertidal zones accurately and time efficiently.

Compact polarimetric (CP) synthetic aperture radar (SAR) can provide ocean surface observations with large-coverage swath and abundant polarimetric scattering information. These distinctive characteristics make CP SAR a potential tool for operational monitoring oil slicks and oil platforms, overcoming the shortcomings of small spatial coverage by the traditional quad-polarization (quad-pol) SAR. In this paper, we use the RADARSAT-2 C-band quad-pol SAR data to generate CP covariance matrix elements and subsequently construct pseudoquad-pol backscatter coefficients, using two CP reconstruction algorithms to evaluate CP SAR's applications in detection of oil slicks and oil platforms. The reconstructed co- and cross-polarization data show good agreement with original radar observations acquired at different incidence angles and wind speeds. Furthermore, we develop an unsupervised classification method using the relative phase, a logical scalar threshold that separates odd and multiple scattering events, as an indicator to discriminate oil slicks and platforms from clean ocean waters. The relative phase is positive for clean ocean surfaces where odd scattering is dominant, but negative for oil platforms and oil slick-covered areas associated with multiple scattering. The detections of oil spills and oil platforms are validated against known oil platform geographic positions and optical aircraft surveys of the oil slicks. The proposed method provides a promising technique to detect oil slicks and oil platforms from CP imaging mode SAR data, i.e., as may be acquired by the RISAT-1, ALOS-2, and the future RADARSAT Constellation Mission.

A hybrid backscattering model is built to provide a consistent description for C-band VH- and VV-polarized normalized radar cross sections (NRCSs). Ocean surface coand cross-polarized NRCS are both treated as a sum of Bragg and non-Bragg scattering components. To better understand the synthetic aperture radar (SAR) observed NRCS signals under high-wind conditions, five C-band RADARSAT-2 dual-polarization SAR hurricane images and the collocated wind vectors measured by the airborne stepped-frequency microwave radiometer (SFMR) are collected. Based on the match-up data, we add a non-Bragg term in the composite Bragg theory to explain the discrepancy between the measurements in the cross-polarization channel and the existing theory results. The non-Bragg scattering to Bragg scattering ratio (Br) is found to be a constant. We build the hybrid backscattering model with Br and establish a relationship between the cross-polarization NRCS and the radar incidence angle under different wind conditions. The NRCS dependence on incidence angle is simulated by the hybrid backscattering model. Finally, a C-band Cross-Polarization Coupled-Parameters Ocean (C-3PO) model is developed to retrieve hurricane winds using VH-polarized ScanSAR by including the radar incidence angle. The collocated SAR and SFMR data sets are separated into two parts: data set-A, for hybrid backscattering model derivation and C-3PO model coefficients tuning, and data set-B, for hurricane wind validation. C-3PO model validation results show that the model is suitable for ocean surface wind mapping from RADARSAT-2 cross-polarization ScanSAR images. The retrieval has a rootmean-square error less than 3 m/s for wind speed up to 40 m/s.

Over the last decades, data from spaceborne synthetic aperture radar (SAR) have been used in hurricane research. However, some issues remain. When wind is at hurricane strength, the wind speed retrievals from single-polarization SAR may have errors, because the backscatter signal may experience saturation and become double valued. By comparison, wind direction retrievals from cross-polarization SAR are not possible until now. In this paper, we develop a 2-D model, the symmetric hurricane estimates for wind (SHEW) model, and combine it with the modified inflow angle model to detect hurricane morphology and estimate the wind vector field imaged by cross-polarization SAR. By fitting SHEW to the SAR derived hurricane wind speed, we find the initial closest elliptical-symmetrical wind speed fields, hurricane center location, major and minor axes, the azimuthal (orientation) angle relative to the reference ellipse, and maximum wind speed. This set of hurricane morphology parameters, along with the speed of hurricane motion, are input to the inflow angle model, modified with an ellipse-shaped eye, to derive the hurricane wind direction. A total of 14 RADARSAT-2 ScanSAR images are employed to tune the combined model. Two SAR images acquired over Hurricane Arthur (2014) and Hurricane Earl (2010) are used to validate this model. Comparisons between the modeled surface wind vector and measurements from airborne stepped-frequency microwave radiometer and dropwindsondes show excellent agreement. The proposed method works well in areas with significant radar attenuation by precipitation.

Too many terrains of coastal zone require effective management to better understand coastal changes. With the development of full-polarization Synthetic Aperture Radar (POL-SAR) imaging, these terrains from coastal zone classification are available. But the classification of coastal land-use types based on full-polarization SAR data has not been investigated. In fact, the signal return coming from the sea can be frequently indistinct from one coming from the land. And coastal zone terrains modeled as a strong multiplicative noise as well as they are very similar with scattering mechanism, which makes the coastline zone types classification a very complicated issue. A joint sparse representation-based method for classification of coastal land-use types with full-polarization SAR data is proposed in this paper.

The robust wind wave growth functions established with ideal fetch-limited and quasi-steady wind conditions (e.g., [11-13]; and references therein) have been observed to be applicable to wind generated waves under considerably more varying conditions, including hurricanes [1, 3, 4, 14, 15] and rapidly accelerating and decelerating wind fields such as those encountered in mountain gap winds [16-19].

Ocean surface wind retrieval from Sentinel-1SAR data is scientific and technically matured at NOAA. The SAR wind has a fairly good accuracy with a standard deviation of less than 2 m/s.

Five possible mechanisms for the rain effects on the spaceborne C-band SAR observations are considered: 1) attenuation and 2) volume backscattering for the microwave transfer in atmosphere; as well as 3) rain - induced damping to the wind waves and 4) rain - generated ring waves on the ocean surface, and 5) diffraction on the sharp edges of rain products. A composite radar scattering model composed of the atmosphere radiative transfer model and the ocean surface Bragg wave theory is employed to analyze the impact of rain on the normalized radar-backscatter cross-section (NRCS) measured in the VV- and cross-polarized C-band Synthetic Aperture Radar (SAR) channels. Our composite model was validated with the matchup of observations of wind speed and rain rate from the SFMR data as well as the NRCSs and incidence angles of C-band VV-polarization SAR data over two hurricanes. Comparisons between the observed NRCSs and the atmosphere part of our model imply that the most important mechanism for the rain effect on the C-band VV polarized SAR hurricane observations is through the influence of waves on the ocean surface. Moreover, the non-Bragg scattering is important for the cross-polarized NRCS simulations.

SST cooling episodes in the early Spring are often observed by various remote sensing measurement systems near the coast of Qingdao, a city in China's Shandong province (Figure 1). Several wind-SST interaction mechanisms that can cause colder deep water to force its way upward and drive away and subsequently replace the warmer surface water, also known as upwelling, have been previously documented. Most of the previous coastal upwelling work has been focused on the mechanism of upwelling processes on a large scale and caused by winds blowing parallel to the coast and Ekman transport. Upwelling on small local scales such as the one shown in Figure 1 has not been extensively studied. However, on a more local and smaller scale, not associated with Coriolis force or Ekman transport, winds blowing offshore can also push water mechanically away from land to produce upper level divergence and upwelling. But it is not clear if the recurring upwelling in Figure 1 can be attributed to the along-coast winds or the offshore winds or the local bathymetry, or a combination of several factors. We will look into the possible mechanisms.

In this work, an automatic method is proposed to determine the center of tropical cyclones (TCs) from a series of RADARSAT-1 synthetic aperture radar (SAR) images, which captured TCs over the Atlantic, the Pacific and the Indian Ocean during the years from 2001 to 2007. The TC centers determined by the method are compared to the TC best track (BT) datasets provided by National Hurricane Center (NHC) of National Oceanic and Atmospheric Administration (NOAA) and Shanghai Typhoon Institute (STI) of the China Meteorological Administration (CMA). The results show a good agreement between the SAR-estimated TC center positions and the BT data, indicating that satellite SAR is a powerful tool for the study of tropical cyclone morphology and dynamics.

In this study, an atmospheric phenomenon, land breeze system, was observed on one image mode high resolution SAR image and one middle resolution optical image. WRF model was implemented with actual meteorological conditions as inputs to successfully simulate this process with some understandable discrepancies. The overpass time of ASAR and MODIS are at 10:35 AM and 11:10 AM local time in Hainan. Long time in-situ observations show that the land breeze process is converting to sea process between 10:00 AM and 12:00 AM. Thus, the satellite observed land breeze front is at its ending stage. That is one possible reason that land breeze front on ASAR image is clearer than on MODIS image. It also explained why the cloud line in MODIS image cannot maintain its shape and started to dissipate. The MODIS image also shows that the land breeze dissipation process is not spatial uniform along the whole land-sea breeze front.

Compact polarimetric (CP) synthetic aperture radar (SAR) is a potential tool for operationally monitoring oil slicks and oil platforms because its large-coverage swath and abundant polarimetric scattering information. In this study, we use C-band RADARSAT-2 quad-polarization (quad-pol) SAR data to simulate CP covariance matrix elements and then construct pseudo quad-pol scattering coefficients by utilizing two different CP reconstruction algorithms. We develop an unsupervised classification method to discriminate oil slicks and platforms from clean ocean waters, using the relative phase, a logical scalar threshold that separates odd and even scattering events. The relative phases are estimated with the reconstructed co- and cross-polarization backscatters, which are positive in clean ocean surfaces where odd scattering is dominant but are negative for oil platforms and oil slick-covered areas associated with even scattering.

The Bragg scattering theory does not explain the cross-polarized (VH or HV) normalized radar-backscatter cross-sections (NRCSs) from the ocean surface very well. The difference lies in the non-Bragg scatterings mechanisms including specular reflection and quasi-specular reflection or diffraction on the sharp edges of wave breaking. The specular reflection is a function of radar incidence angle and is negligible for the intermediate incidence angle. The quasi-specular reflection and diffraction on the sharp edges of wave breaking is a function of incidence angle, wind speed and direction. In this study, we adopt the C-band Cross-polarization Ocean (C-2PO) model as the total backscatter from the ocean surface, and assume the non-Bragg scattering as the differences between NRCSs simulated by C-2PO model and composite Bragg model. And a non-Bragg scattering ratio is suggested. With the collocated datasets of wind speeds, wind directions, measured by airborne SFMR and NRCSs from SAR, the suggested non-Bragg scattering ratio is validated.

Classifying different types of land cover in coastal zones using synthetic aperture radar (SAR) imagery is a challenge due to the fact that many types of coastal zone have similar backscattering characteristics. In this letter, we propose an unsupervised method based on a three-channel joint sparse representation (SR) classification with fully polarimetric SAR (PolSAR) data. The proposed method utilizes both texture and polarimetric feature information extracted from the HH, HV, and VV channels of a SAR image. The texture features are extracted by applying a wavelet transform to a SAR image, and then sparsely represented based on the correlation among the three channels. The polarimetric features, i.e., the scattering entropy and scattering angle from the H/α model, are also sparsely represented. A joint SR algorithm using both texture and polarimetric features is constructed to establish target dictionaries. An orthogonal matching pursuit algorithm is then used to calculate sparse coefficients. Hybrid coefficients are inputted to the kernel support vector machine for a fully PolSAR image classification. We applied the proposed algorithm to an Advanced Land Observing Satellite-2 L-band SAR image acquired in the Yellow River Delta, China. The classified land types are validated against the official survey map. The algorithm performs well in distinguishing six coastal land-use types. A comparison study is also conducted to show that proposed algorithm outperforms two commonly used classification methods.

In this paper, a theoretical and experimental analysis of polarimetric synthetic aperture radar (SAR) architectures is undertaken for sea oil slick observation purposes. Reference is made to the conventional full-polarimetric (FP) SAR that is here contrasted with new-generation polarimetric SAR architectures, known as compact-polarimetric (CP) SAR. Two CP modes are considered, i.e., the hybrid-polarity and π/4 modes, whose measurements are emulated from actual L- and C-band FP SAR data. Polarimetric sea surface scattering is predicted according to an extended version of the Bragg scattering model (X-Bragg) in order to point out the differences exhibited between FP and CP SAR architectures and among CP SAR modes. Theoretical predictions are then contrasted with experiments undertaken on actual polarimetric SAR data collected over well-known oil slicks and weak-damping surfactants. Results confirm model prediction, showing that differences mainly apply when polarimetric features are estimated over slick-free sea surface using different SAR architectures, with the π/4 mode behaving closer to FP SAR. Although CP SAR architectures measure only a subset of the FP information content, they represent an interesting operational alternative for both detecting oil slicks and discriminating them from weak-damping surfactants.

Two cases of mode-2 internal solitary waves (ISWs) induced by an anticyclonic eddy (AE) were clearly present in two synthetic aperture radar images acquired in the South China Sea (SCS) in April 2001. Similar ISWs were repeatedly observed in the same area one month later in May, but in that case, the ISW patterns indicated that they were regular mode-1 ISWs. To confirm that those ISWs in April are of mode-2 type, we analyze the in situ and other remote sensing data and propose two possible mode-2 ISW generation processes: 1) an eddy-induced change of the water stratification, which results in favorable hydrographic conditions for internal wave generation, and 2) the resonance between mode-1 internal tides and AE excites mode-2 internal tides, and the mode-2 internal tides disintegrate into mode-2 ISWs. The sea level anomaly and sea surface temperature data in April 2001 show that an AE existed in this area in April. Argo profile data within AEs from 1990 to 2014 in the SCS are used to show how an AE affects the vertical water properties and creates favorable conditions for mode-2 ISWs, and these conditions did exist in the study area in April 2001. The observed mode-2 ISW cases provide the first evidence of the existence of eddy-induced mode-2 ISWs in the SCS.

We present a method to retrieve wind speeds in hurricanes from spaceborne passive microwave radiometer data. Brightness temperature (TB) observations acquired at the 6.9-GHz horizontal polarization channel by the AMSR-E and AMSR2 onboard the Earth Observing System Aqua and Global Change Observation Mission-Water 1 satellites are selected for wind retrieval due to the fact that the signal at this frequency is sensitive to high wind speeds but less sensitive to rain scatter than those acquired at other higher frequency channels. The AMSR-E and AMSR2 observations of 53 hurricanes between 2002 and 2014 are collected and collocated with stepped-frequency microwave radiometer (SFMR) measurements. Based on the small slope approximation/small perturbation method model and an ocean surface roughness spectrum, the wind speeds are retrieved from the TBdata and validated against the SFMR measurements. The statistical comparison of the entire data set shows that the bias and root-mean-square error (RMSE) of the retrieved wind speeds are 1.11 and 4.34 m/s, respectively, which suggests that the proposed method can obtain high wind speeds under hurricane conditions. Two case studies show that the wind speed retrieval bias and RMSE are 1.08 and 3.93 m/s for Hurricane Earl and 0.09 and 3.23 m/s for Hurricane Edouard, respectively. The retrieved wind speeds from the AMSR-E and AMSR2 continuous three-day observations clearly show the process of hurricane intensification and weakening.

Several groups of atmospheric gravity waves (AGWs) were observed on a Sentinel-1A synthetic aperture radar (SAR) image acquired near the Kuril Islands in the Northwest Pacific Ocean on June 1, 2015 during the passage of an extratropical cyclone (ETC). These waves occurred on the lee side of the mountains located on the islands. Both diverging and transverse waves with wavelengths ranging between 20 and 30 km are shown as alternating bright-dark patterns in the SAR image. For the diverging waves, there exists a prominent asymmetry in the wave motions of the two arms. The Moderate Resolution Imaging Spectroradiometer and Landsat 7 Enhanced Thematic Mapper Plus images acquired 5-7 h prior to the Sentinel-1A pass also contain the same groups of AGWs. The mesoscale Weather Research and Forecasting model simulation confirms that the AGWs are lee waves triggered by the airflow over the islands. AGWs are aligned perpendicular to the wind direction and locked on the lee side of the islands. The life span of the waves is about two days, consistent with that of the ETC over the region. The numerical model also successfully reproduces the main characteristics of the lee waves. Simulation results demonstrate that the variation in the wave parameters (i.e., wavelength, amplitude, orientation, wedge angle of the diverging wave, and vertical propagation characteristic) and the wave asymmetry of the diverging wave are mainly caused by the wind and stratification changes. The smaller amplitude of the diverging wave seems to be associated with a smaller Froude number.

This paper compares the typhoon centers from the tropical cyclone best track (BT) data sets of three meteorological agencies and those from synthetic aperture radar (SAR) and infrared (IR) images. First, we carried out algorithm comparison using two newly developed algorithms and one existing wavelet-based algorithm, which were used to extract typhoon eyes in six SAR images and two IR images. These case studies showed that the extracted eyes by the three algorithms are consistent with each other. The differences among them are relatively small. However, there is a systematic difference between the extracted centers and the typhoon centers from the three BT data sets, which were interpolated to the imaging times first. We then compared the typhoon centers determined from 25 SAR and 43 IR images with those from the three BT data sets to investigate the performance of the latter at the sea surface and at the cloud top, respectively. We found that the typhoon centers from the three BT data sets are generally closer to the locations extracted from the SAR images showing sea-surface imprints of the typhoons than those from the IR images showing cloud-top structures of the typhoons.

In this paper, we present the synthetic aperture radar (SAR) observation of the detailed sea surface wind patterns associated with fetch in the Bohai Sea, China. We then implemented the WRF model to simulate the entire processes of this weather event. WRF model results show the dynamics and evolution of this event.

The computation of wind speeds at high (<; 1 km) resolution from spaceborne synthetic aperture radar (SAR) is a mature geophysical application. A number of researchers, a modest sample of which are cited [1, 2, 3, 4, 5, 6, 7], have described the geophysical relationship between normalized radar cross section (NRCS) and ocean surface wind speed, and how this relationship can be exploited to infer wind speeds. In May 2013, NOAA began the operational production of SAR-derived wind speed maps using Canadian Radarsat-2 imagery purchased by the U.S. National Ice Center. Monaldo et al. [8] chronicle the history of the first observation of ocean wind features from the Seasat SAR and its evolution to an operational product.

After many years of development and experimental production, the United States National Oceanic and Atmospheric Administration (NOAA) transitioned the generation of winds derived from satellite synthetic aperture radar (SAR) data to operational status on May 1, 2013, employing SAR data from the Canadian RADARSAT-2 satellite. Winds of 500 m resolution are being produced from SAR data being purchased for the U.S. National Ice Center. Products are distributed internally within NOAA and via public websites. Currently the production of winds from Sentinel-1A data are being added to this system along with the ability to produce new product output formats (e.g., CoastWatch) and a wind archive is being developed by the NOAA National Centers for Environmental Information.

Mode 2 internal wave (IW) signatures are observed by the RADARSAT-1 synthetic aperture radar (SAR) in the Northeast South China Sea on April 07, 2001. In this study, we prove that the conditions of the study area are favorable for the generation and propagation of mode-2 waves to appear in SAR imagery. The generation mechanism of this mode 2 IW is mode-1 IW evolves into mode-2 convex IW packets when shoaling.

Synthetic aperture radar (SAR) images reveal the surface imprints of atmospheric occluded fronts. An occluded front is characterized as a low-wind zone located between and within two zones of higher winds blowing in the opposite directions on the left and right sides of the occluded front. A group of four SAR images reveal that the width of an individual occluded frontal zone and the wind magnitudes outside fronts vary greatly from case to case. In this paper, we performed a case study to analyze an occluded front observed by an Environmental Satellite (Envisat) Advanced SAR and ASCAT scatterometer along the west coast of Canada on November 24, 2011. The two-way interactive, triply nested grid (9-3-1 km) Weather Research and Forecasting (WRF) model was utilized to simulate the evolution of the occluded front. The occluded front moved toward the east during a 24-h model simulation, and the movement between 18:00 and 21:00 UTC matched the occluded front positions derived from the concurrently collected surface weather maps; from the National Oceanic and Atmospheric National Weather Service archives. The WRF-simulated low-wind zone associated with the occluded front and ocean surface wind speed match well with the SAR and scatterometer wind retrievals. High wind outside the front zone became weaker during the front evolution, whereas the width of the occluded frontal zone was contracted laterally. Analysis of the WRF model derived potential temperature field suggests that the occlusion process occurred below the 800-mb level. The structure of the occluded front studied here not only follows the conventional conceptual model and also supports the findings of a novel wrap-up conceptual model for an atmospheric frontal occlusion process.

Spaceborne high-resolution synthetic aperture radar (SAR) is a potential powerful tool for rainfall pattern and intensity observations over the sea surface. However, many interesting rain-related phenomena revealed by SAR images are still not fully understood due to poor theoretical modeling of the rain–wind–wave interactions. This paper attempts to develop a physics-based radiative transfer model to capture the scattering behavior of rainfall over a rough sea surface. Raindrops are modeled as Rayleigh scattering nonspherical particles, whereas the rain-induced rough surface is described by the Log-Gaussian ring-wave spectrum. The model is validated against both empirical models and measurements. A case study of collocated Envisat ASAR data and NEXRAD rain data is presented to demonstrate the performance of the newly developed model. Finally, numerical simulation results suggest that rain-related scattering becomes significant as compared with wind-related scattering when the frequency is above C-band, whereas the raindrop volumetric scattering becomes significant above X-band.

The multiscale normalized cut technique is used to detect land/sea boundary in synthetic aperture radar (SAR) images. The proposed method is a region-based one, which is able to capture and utilize spatial information in the image. The real SAR images, e.g. ERS, ENVISAT and COSMO-SkyMed (CSK) SAR images, together with the Landsat-7 ETM+ optical data and in-situ GPS data were collected and used to validate the performance of the proposed method for coastline detection in SAR images.

The multiscale normalized cut technique is used to detect oil spill in synthetic aperture radar (SAR) images. The proposed approach is a region-based one, which is able to capture and utilize spatial information in the image. Actual ENVISAT ASAR image was collected and used to validate the effectiveness and efficiency of the proposed approach for oil spill detection in SAR images.

In this paper, the Ulva prolifera bloom event in the Yellow Sea in summer 2010 is investigated by MODIS (Moderate Resolution Imaging Spectroradiometer) images. We use the FAI (Floating Algae Index) method to detect the distribution of the floating macroalgae from the images. Then we apply the GNOME (General NOAA Operational Modeling Environment) model to simulate the trajectories of the Ulva prolifera in the Yellow Sea. The model results agree well with satellite observations, indicating that the occurrence and movement of the floating macroalgae can be investigated with the combination of GNOME model and satellite data.

The spatial and temporal variability of internal solitary waves (ISWs) speed around Dongsha Atoll in the South China Sea (SCS) was investigated using multi-satellite image pairs separated by about 0.5-2.5 hours. SAR image pair, VIIRS/MODIS and SAR/MODIS pairs in July 2007, March 2009, and May 2013 were analyzed, respectively. The ISW phase speeds were derived using the horizontal displacement of the ISW patterns and the time difference between the 2 satellite images. The phase speeds were in good agreement with the theoretical calculations using the Taylor-Goldstein (T-G) equation with a non-linear term. The ISW phase speed decreases from east to west and from south to north. The temporal variability of ISW phase speeds are mainly affected by water depth, with minor seasonal variations.

In this work we investigate the deformation signals affecting the coastal region of the megacity of Shangai (China) where, to satisfy the growing land demand for industrial and urban development, man-made lands reclaimed from the sea have been retrieved and used to build airports, harbors, and industrial areas. The presented analysis is carried out through the application of spaceborne Differential Synthetic Aperture Radar (SAR) Interferometry (DInSAR) techniques. A general picture of the ground deformation field is provided by applying the Small Baseline Subset (SBAS) as well as the Permanent Scattereres (PS) approaches to archives of SAR images collected by ENVISAT and COSMO-SkyMed sensors from 2007 to 2014 over the investigated area.

A two-step physically consistent procedure is proposed to exploit COSMO-SkyMed (CSK) synthetic aperture radar (SAR) data acquired in the incoherent dual-polarization PingPong mode. The first step, which deals with land/sea discrimination, is accomplished in a robust and effective way by exploiting the inherent peculiarities of the CSK PingPong mode. Hence, a dual-polarization scattering model that relates the correlation between the HH and VV CSK polarimetric channels to the coherence time of the observed scene is first proposed. The second step, which deals with the extraction of the continuous coastline, is accomplished by a simple image processing that consists of extracting intermediate frequency components using two Gaussian-shaped filters. Experiments undertaken over actual CSK single-look slant range complex PingPong HH/VV SAR data show the physical soundness of the proposed rationale and the processing effectiveness.

Automatic hurricane information extraction in synthetic aperture radar (SAR) images has been a research topic in development. In this study, using saliency-based visual attention model, we developed an image processing procedure to extract hurricane eyes from SAR images. Experiment results show that hurricane eyes can be well extracted even when it is not visually obvious in images.

Tropical cyclones generate powerful wind, torrential rainfall, high waves and damaging storm surge that affect coastal communities. Tracking and predicting cyclones is one of the most important tasks for meteorologists. In this study, we compare the hurricane/typhoon eye locations at the sea level observed by spaceborne Synthetic Aperture Radar (SAR) and its counterpart at the cloud level by the simultaneous infrared imagery. The vertical eye tilt at these two heights is compared with 850-200hPa vertical wind shear from SHIPS data. Five case studies show that the displacements vary from 10 to 22 km, with tilt direction oriented from downshear-left to downshear to downshear-right. These results are consistent with former studies.

In this study, a comparison of wind field measurements from HY-2A Scatterometer and U.S. National Data Buoy Center (NDBC)'s moored buoys is performed. These comparisons were made in Pacific Ocean and Atlantic Ocean over one month period of August 2012. The SCAT wind speed retrieval agreed well with the buoy measurements, with mean differences of -0.74 m/s and standard deviations of 1.52 m/s. The results indicate that SCAT-derived ocean surface wind speeds are as accurate as other scatterometer, such as Quikscat and ASCAT. However, the wind direction retrieval still has some problems need to be investigated in the future.

We present a innovative physically-based approach to detect oil platforms by using compact polarimetric (CP) synthetic aperture data. We utilize two RADARSAT-2 fine quad-polarization SAR images of Gulf of Mexico to estimate the CP covariance matrix. The resulting CP data is used to reconstruct pseudo quad-polarization covariance matrix. The Stokes parameters are evaluated using the reconstructed pseudo quad-polarization covariance matrix elements. We further calculate the relative phase with two Stokes parameters. The relative phase is a logical scalar descriptor which can be used to detect the oil platforms. The detected oil platforms are validated with ground truth from NOAA. The proposed method provides a simple and effective mapping technique for oil platform detection.

In this paper, we performed a comparison of wind speed from synthetic aperture radar (SAR), scatterometer, moored buoys and numerical model. These comparisons were made in near U.S. coast regions. The results indicate that SAR-derived ocean surface wind speeds are as accurate as the scatterometer and model wind products.

In this study, we analyze 83 synthetic aperture radar (SAR) images including 73 from RADARSAT-1 and 10 from ENVISAT that contain tropical cyclone eye information. We also obtain ancillary tropical cyclone intensity information from NOAA National Hurricane Center and Japan Meteorological Agency. Based on this information, we generate tropical cyclone morphology statistics. We found that majority of the hurricanes are in wavenumber 1 and 2 category. Marine atmospheric boundary layer rolls can also be extracted from SAR image.

COSMO-SkyMed (CSK) dual-polarized Synthetic Aperture Radar (SAR) data, acquired in PingPong mode, are first exploited to extract coastline. A dual-polarization model that takes full benefit of the PingPong mode peculiarities is developed to distinguish sea surface from land. Following this rationale, a physically-based filtering technique is proposed to extract the coastline in full-resolution CSK SAR data. Experiments confirm the soundness of the proposed approach and its effectiveness in terms of processing time.

The relationship between ocean surface dark/bright pattern of internal wave in SAR images and sea surface wind field is investigated. Two cases of satellite SAR images have ISWs are analyzed, and SAR signature of ISWs under different current and wind conditions are simulated. It is shown that the sequence of brighter and darker stripes of ISWs signatures in SAR images are relative to surface current. The wind direction and wind speed differences can affect the shape and brightness of ISWs stripes, but cannot change the sequence of the stripes.

We present an efficient algorithm for retrieving the ocean-surface wind vector from C-band Radar Satellite RADARSAT-2 fully polarimetric synthetic aperture radar (SAR) measurements based upon the copolarized geophysical model function, i.e., CMOD5.N, and the cross-polarized ocean backscatter model, i.e., C-2PO. The analysis of fine quad-polarization mode single-look complex SAR data and collocated in situ moored buoy observations reveals that the polarimetric correlation coefficient between co- and cross-polarization channels has odd symmetry with respect to the wind direction. This characteristic is different from the feature that normalized radar cross sections for quad-polarization have even symmetry regarding the wind direction. We first use the C-2PO model to directly retrieve wind speeds without any external wind-direction and radar-incidence-angle inputs. Subsequently, the retrieved wind speeds, along with incidence angles and CMOD5.N, are employed to invert the wind direction, still with ambiguities. The odd-symmetry property is then applied to remove the wind direction ambiguities. Thus, it is shown that fully polarimetric SAR measurements provide complementary directional information for the ocean-surface wind fields. This method has the potential to improve wind vector retrievals from space.

Marine atmospheric boundary phenomena, i.e., atmospheric gravity waves (AGW), vortex streets and boundary rolls, modulate the surface wind field and will leave imprints visible on synthetic aperture radar (SAR) images. In this case study, we present an ENVISAT SAR observation of AGW offshore of the mountain Laoshan along the yellow sea coast of China. The Weather Research and Forecasting (WRF) weather model is used to simulate the development of this group of AGW. The model simulation and SAR observation agree reasonable well.

SAR-derived wind measurements are in the process of being implemented for operational production within NOAA's National Environmental Satellite, Data, and Information Service. For C-band ENVISAT and RADARSAT-1/2 data, the CMOD5 algorithm is being used; for ALOS data, a special L-band wind algorithm is employed. Comparisons of both C-band and L-band winds with ASCAT scatterometer wind measurements show biases of 0.58 m/s or less and standard deviations of 1.31 m/s or less. SAR wind vectors will be stored in a NetCDF4-formatted file and made available in a number of product formats via the NOAA CoastWatch program.

High spatial resolutions synthetic aperture radar (SAR) retrieved ocean surface wind field under actual conditions can be affected by several ocean surface features, such as artificial object, surface oil slicks and air-sea boundary layer stabilities etc. In this paper we present several case studies of those impacts, and a semi-empirical model to correct the air-sea boundary layer stabilities. We demonstrate that the new model helps to improve the wind retrieval accuracy in the Gulf Stream north wall areas.

In this letter, we generate a temporal/spatial matchup data set between QuikSCAT scatterometer and RADARSAT-1 synthetic aperture radar (SAR) wind products in offshore waters along the U.S. West Coast. Analysis of the resulting three-year database shows that, in general, the wind products from both sensors have characteristics similar to those reported in the literature. Then, we perform an error analysis in the space domain and find that there is significant discrepancy between the two wind products as the matchup points move closer to the coast. The root-mean-square error (rmse) and standard deviation (STD) between the two data sets increases markedly for points matched within about 100 km of the coastline. Beyond 100 km, the rmse, STD, and systematic bias become small and stable. In addition, an empirical relationship between QuikSCAT and SAR winds in coastal region is proposed. Thus, the bias and errors should be taken into account if the standard operational QuikSCAT wind products are used for forcing models in the coastal ocean.

Within the National Environmental Satellite, Data, and Information Service, National Oceanic and Atmospheric Administration, multiplatform synthetic aperture radar (SAR) imagery is being used to aid post hurricane and postaccident response efforts in the Gulf of Mexico, such as in the case of the recent Deepwater Horizon oil spill. The main areas of interest related to such disasters are the following: (1) to identify oil pipeline leaks and other oil spills at sea and (2) to detect man-made metallic targets over the sea. Within the context of disaster monitoring and response, an innovative processing chain is proposed to observe oil fields (i.e., oil spills and man-made metallic targets) using both Land C-band full-resolution and fully polarimetric SAR data. The processing chain consists of two steps. The first one, based on the standard deviation of the phase difference between the copolarized channels, allows oil monitoring. The second one, based on the different symmetry properties that characterize man made metallic targets and natural distributed ones, allows man made metallic target observation. Experiments, accomplished over single-look complex L-band Advanced Land Observing Satellite (ALOS) Phased Array type L-band Synthetic Aperture Radar (PALSAR) and C-band RADARSAT-2 fully polarimetric SAR data gathered in the Gulf of Mexico and related to the Deepwater Horizon accident, show the effectiveness of the proposed approach. Furthermore, the proposed approach, being able to process both Land C-band fully polarimetric and full resolution SAR measurements, can take full benefit of both the ALOS PALSAR and RADARSAT-2 missions, and therefore, it allows enhancing the revisit time and coverage which are very critical issues in oil field observation.

In this paper, we perform a comparison of wind speed measurements from the ENVISAT Advanced Synthetic Aperture Radar (ASAR), the MetOp-A Advanced Scatterometer (ASCAT), the U.S. National Data Buoy Center's moored buoys, and the U.S. Navy Operational Global Atmospheric Prediction System (NOGAPS) model. These comparisons were made in near U.S. coast regions over a 17-month period from March 2009 to July 2010. The ASAR wind speed retrieval agreed well with the scatterometer and model estimates, with mean differences ranging from -0.69 to 0.85 m/s and standard deviations between 1.16 and 1.77 m/s, depending upon the ASAR beam mode type. The results indicate that ASAR-derived ocean surface wind speeds are as accurate as the ASCAT and NOGAPS wind products. Comparisons between ASCAT winds and synthetic aperture radar (SAR) winds averaged at different spatial resolutions show very little change. This demonstrates that it is suitable that the scatterometer wind retrieval geophysical model function, i.e., CMOD5, is used for SAR wind retrieval. The impact of C-band VV polarization SAR calibration error on wind retrieval is also discussed.

Synthetic aperture radar (SAR) observes the large-scale ocean surface wind field. With SAR instruments, we can actively monitor phenomena in the coastal ocean and marine atmospheric boundary layer at very high spatial resolution (on the order of tens of meters) in all weather conditions day and night. SAR observations are particularly useful in coastal regions where clouds are usually present, causing observation problems for visible and infrared sensors. SAR sensors onboard the RADARSAT-1/2, ENVISAT, ALOS, and other satellites can provide swath coverage of about 100 to 450 km, wide enough to cover oceanic and atmospheric meso-scale features. SAR has long been used to monitor the ocean surface wind field, vessel locations, oil spills, sea state, and sea ice at NOAA. In this paper, we present several case studies.

In this study, two innovative physically-based approaches have been developed to detect man-made metallic objects and oil slicks in polarimetric SAR data. They are based on the different sea surface scattering mechanisms expected with and without oil slicks and metallic objects. Experiments, accomplished over Single Look Complex (SLC) Level 1.1 quad-pol L-band ALOS PALSAR SAR data, demonstrate the effectiveness of the two approaches for oil slick and metallic target detection purposes and witness the capability of ALOS PALSAR data for such applications.

A group of submerged ocean bottom sand ridges in the Bohai Sea, China, are shown in RADARSAT-1 and ENVISAT synthetic aperture radar (SAR) images. The sand ridges appear as fingerlike quasi-linear features in the SAR images. Examining the detailed local bathymetry chart, we find that these features coincide with the satellite images. The heights of the sand ridges are less than 10 m, and the water depth is between 10 and 30 m. The spacing of the sand ridges is about 10 km, and the length of the sand ridges is about 20 km. The same sand ridges are also visible on a Moderate Resolution Imaging Spectroradiometer (MODIS) true-color image. The semidiurnal and diurnal tidal currents in this area are almost parallel to the major axis of these sand ridges. These observations cannot be explained using the existing 1-D SAR imaging model, which is not applicable to sand ridges parallel to the tidal current. In this paper, we consider the shallow-water current bathymetry in a 2-D space. An analytical ocean model was applied to demonstrate the temporal variations of the current divergence and convergence that are induced by the along-sand-ridge-direction current and ridge interaction. A radar simulation model is used to simulate the variation of normalized radar cross section (NRCS) induced by the ocean surface current. The simulated NRCS variation is similar to that extracted from the calibrated SAR image. Simulation results also show that the NRCS variation becomes negligible when the ocean current is set to about half of the maximum tidal current.

The change in accuracy of synthetic aperture radar (SAR) near-surface wind measurements as a function of increasing wind speed is assessed by matching satellite SAR measurements with wintertime buoy measurements in the Bering Sea. For wind speeds less than 15 m/s, SAR-buoy wind speed comparisons show biases less than 1 m/s with standard deviations less than 2.5 m/s. For wind speeds from 15 to 25 m/s, SAR wind measurements have a positive bias above 2 m/s and somewhat more scatter when compared to buoy winds. For winds larger than 25 m/s, the accuracy of SAR winds is not well known. Very few buoy matches are available. Comparisons of SAR winds with hurricane model winds and aircraft measurements are still in a preliminary stage, but indicate that improved algorithms are needed.

Two projects undertaken by the National Oceanic and Atmospheric Administration (NOAA) National Environmental Satellite, Data, and Information Service (NESDIS) have shown success in using spaceborne synthetic aperture radar (SAR) to identify oceanic eddies and current boundaries. In addition to detecting the frontal area and change in slick patterns, the SAR imagery may also pick up a change in the low-level wind structure as a result of the sea surface temperature (SST) gradient between the eddy and its surroundings affecting the marine atmospheric boundary layer (MABL) stability. This wind fluctuation modulates the sea surface roughness, allowing the eddy boundaries to be imaged by SAR. Two examples, one of an eddy in the Gulf of Alaska, and the second of the Loop Current boundary in the Gulf of Mexico, are analyzed to show the correlation between the SST and surface wind gradients across their boundaries.

In this study, the sea surface imprints of strong mountain katabatic winds and gap winds are observed on RADARSAT-1 synthetic aperture radar (SAR) ScanSAR wide images off the west coast of the U.S. and in the Gulf of Alaska. Two case studies are presented. In the first case study, a RARASAT-1 SAR scene taken at 14:25:30 UTC on January 21, 2003 shows a finger-like wind pattern that mirrors the coastal mountain height. In the second case, the SAR image was taken at 4:41:45 UTC on December 22, 1999. It shows a strong gap wind and vortex streets through the Aleutian Islands. In order to understand the dynamics of these wind patterns observed in the SAR images, we simulated the low level atmospheric circulation using the fifth-generation Pennsylvania State University (PSU)-National Center for Atmospheric Research (NCAR) Mesoscale Model, MM5. A triple nested-grid (9/3/1 km) technique is employed to achieve a multi-scale simulation. In general the MM5 model captures the wind pattern very well and reveals the dynamics of these meso-scale atmospheric phenomena. However, the MM5 did not resolve the vortex shedding due to the model resolution and the complex nature of this phenomenon

High-resolution winds derived from RADARSAT-1 synthetic aperture radar (SAR) images have been produced for the waters around Alaska since 1999. Wind speed images show useful details of many meteorological phenomena of interest in weather analysis and forecasting. These include wakes, gap flows, lee waves, atmospheric fronts, cyclones, and barrier jets.

A group of atmospheric solitons is identified on a RADARSAT-1 synthetic aperture radar (SAR) image and a Moderate Resolution Imaging Spectroradiometer (MODIS) image taken about 4.5 hours later on June 6, 2001. On both images, this group of solitons is showed as dark-bright linear features. The atmospheric solitons are generated at St. Lawrence Island in Bering Sea and propagate against the airflow in the upstream direction. On the first SAR image, there are only three wave crests. However, on the second MODIS image, the wave train propagates further upstream and seven wave crests can be identified. The basic properties of this group of solitons are measured and derived. Radiosonde data show that the Froude number is close to unity. Therefore, theoretically, this phenomenon can be described by the classic "flow over bump" model and analyzed using the classic KdV equation.

The National Oceanic and Atmospheric Administration (NOAA) National Environmental Satellite, Data, and Information Service (NESDIS) is in the third year of the Alaska SAR Demonstration (AKDEMO), an applications project using RADARSAT-1 synthetic aperture radar (SAR) and derived products. The success of this demonstration in providing near real-time SAR data, derived products, and other ancillary data to federal and state agencies, has motivated the development of a similar experimental multi-sensor data fusion system for the Gulf of Mexico and Caribbean region called GoMEx. Unlike the AKDEMO system which focused on near-real time data, this project will begin by using archived data from diverse remote sensing sensors such as RADARSAT-1 SAR, GOES imagers, SeaWiFS, MODIS, AVHRR, scatterometers, as well as data from moored buoy measurements and numerical weather models to study issues and phenomena unique to the region. Other agencies that will contribute and use this system include the National Ocean Service (NOS), Louisiana State University, National Marine Fisheries Service and the University of Maryland. Sample applications of this system are presented including detection of algal blooms, coral reefs off of Belize, oil slicks and rigs off of Louisiana, and fishery applications.

A summary of the interim results of the first two years of the authors' NASA RADARSAT-1 ADRO-2 SAR project is given. The Alaska SAR Demonstration (AKDEMO) is providing winds, vessel positions, and SAR imagery to users in Alaska for evaluation as to their utility to operational government agencies responsible for ocean and weather prediction and fisheries management/enforcement. The AKDEMO applications are maturing and their accuracy has been measured. A new demonstration, the Gulf of Mexico Experiment (GoMEx), is now underway to examine use of SAR data and products in hazardous algal bloom (HAB) and oil spill/seep monitoring. Initial results show correspondence of bloom signatures in ocean color and SAR data in areas of high HAB concentration as measured from ship water samples. In addition to these two applications demonstrations, research is underway in the use of SAR data to study upwelling, river plumes, ocean current boundaries, atmospheric boundary layer processes, and other applications. This research will continue into the third and final year of the ADRO-2 project.

We analyze the refraction of long oceanic waves at the Gulf Stream's north wall off the Florida coast as observed in imagery obtained from the RADARSAT-1 synthetic aperture radar (SAR) during the passage of Hurricane Bonnie on August 25, 1998. The wave spectra are derived from RADARSAT-1 SAR images from both inside and outside the Gulf Stream. From the image spectra, we can determine both the long wave's dominant wavelength and its propagation direction with 180/spl deg/ ambiguity. We find that the wavelength of hurricane-generated ocean waves can exceed 200 m. The calculated dominant wavelength from the SAR image spectra agree very well with in situ measurements made by National Oceanic and Atmospheric Administration National Data Buoy Center buoys. Since the waves mainly propagate toward the continental shelf from the open ocean, we can eliminate the wave propagation ambiguity. We also discuss the velocity-bunching mechanism. We find that in this very long wave case, the RADARSAT-1 SAR wave spectra should not be appreciably affected by the azimuth falloff, and we find that the ocean swell measurements can be considered reliable. We observe that the oceanic long waves change their propagation directions as they leave the Gulf Stream current. A wave-current interaction model is used to simulate the wave refraction at the Gulf Stream boundary. In addition, the wave shoaling effect is discussed. We find that wave refraction is the dominant mechanism at the Gulf Stream boundary for these very long ocean swells, while wave reflection is not a dominant factor. We extract 256-by-256 pixel full-resolution subimages from the SAR image on both sides of the Gulf Stream boundary, and then derive the wave spectra. The SAR-observed swell refraction angles at the Gulf Stream north wall agree reasonably well with those calculated by the wave-current interaction model.