Coral reef is a typical marine ecosystem and has significant implications for protecting marine biodiversity, and maintaining marine ecological balance. Accurate geomorphic information is the base of coral reef conservation, which usually is extracted by high-resolution remote sensing. Recent classification methods for coral reef geomorphology always focus on the extraction of deep spectral and texture features, ignoring the inherent geospatial information of geomorphology and losing the shallow-layer information, which leads to low classification accuracy. This paper proposes a deep learning classification method for coral reef geomorphology, named as GCU-Net which integrates the convolutional attention mechanism and the geo-spatial cognition. Experiments were carried out in North Reef and Zhaoshu Island geomorphology with the Gaofen-2 (GF-2) satellite image. The results demonstrate that the GCU-Net’s accurate classification with an overall accuracy (OA) of 90.46% and 88.92%, respectively, which better extracts the useful information in the shallow-layer, and effectively reduces the omission and misclassification of geomorphic types due to the different spatial positions. Our method exhibits excellent classification performance with an OA improvement of over 7% compared to the comparison method. Therefore, the method proposed in this paper is more effective in obtaining accurate information on coral reef geomorphology and can provide technical support for carrying out large-scale fine monitoring of coral reefs.

Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) photon counting data are widely used in nearshore bathymetry. However, a large number of noisy photons unavoidably exist in the acquired photon data, and photon denoising is necessary to accurately extract the underwater signal photons. Based on the phenomenon that the number of photons increases at the target object, a methodology for detecting ICESat-2 bathymetry photons based on quasi full waveform is proposed. First, the ICESat-2 photon data are split at certain intervals to extract the sea surface and seafloor height photons within each interval. Second, a double-peaked Gaussian function is utilized to fit the photon heights in each interval to determine the sea surface and seafloor heights. Third, sea surface and seafloor height photons for each interval are connected along the track direction to generate sea surface and seafloor datums. Finally, the optimal buffer distance is calculated and the buffer is created to filter the photons, and then, the refraction correction is applied to the seafloor photons. ICESat-2 data from Vieques Island and Saint Croix Island are selected for the experiment and the results are compared with in situ bathymetry data. The results show that the mean absolute error (MAE) of the bathymetry results extracted by the proposed method ranged from 0.16 to 0.30 m and the RMSE ranged from 0.24 to 0.32 m in different areas. Compared with density-based spatial clustering of applications with noise (DBSCAN), ordering points to identify the clustering structure (OPTICS), and adaptive elevation difference thresholding algorithm (AEDTA), the proposed method accurately recognizes the signal photons for different densities of photon data and different complexities of seafloor topography, and the extracted signal photons are complete and continuous, showing excellent robustness. By selecting suitable split spacing along the track direction and histogram separation spacing, the proposed method in this study achieves excellent performance in bathymetry extraction.

Marine oil spills pose a significant risk to the ecological balance and human health. It is crucial to promptly and accurately identify the type of oil spill to facilitate emergency response and inform scientific decisions. Remote sensing technology is at the forefront of current research on oil type identification. This article presented comprehensive research on the systematic identification of oil types. The optical and thermal infrared data were gathered for various typical oils to elucidate their optical and thermal characteristics (OTC). On this basis, we developed the oil-type OTC dual-driven identification model (OTC-DDIM). This model incorporates a sample expansion module [OTC-conditional generative adversarial network (CGAN)] to increase sample diversity, a characteristic extraction module (OTC-EM) to extract OTC, and an adaptive identification module to fuse and enhance OTC for identifying oil-spill types. Further research revealed the critical role of optical characteristic screening in eliminating redundant information interference and improving the identification accuracy and efficiency. Temperature, a dominant environmental factor (EF), played a key constraint on the generation of high-quality thermal infrared extension samples by OTC-CGAN. Under ideal oil-spill scenarios, the model demonstrated excellent identification capabilities, achieving an overall accuracy (OA) of 96.15%, with both Kappa and average$F_{1}$-score reaching 0.96. The method verification and application were conducted under simulated oil-spill scenarios. The experimental results demonstrated that OTC-DDIM could accurately and reliably identify oil-spill types using OTC, achieving accuracies of 91.71%, 0.92, and 0.90, respectively. In summary, this study could provide essential technical support for emergency responses to marine oil-spill accidents.

Marine oil spills and emulsions are highly detrimental to the ecological environment and human health. Optical and thermal infrared remote sensing technologies provide important approaches for marine oil-spill detection, but accurate cross-spatiotemporal remote sensing detection in complex scenarios is still challenging. In this article, optical and thermal infrared observation experiments on oil spills and emulsions were designed and conducted to analyze the optical and thermal response characteristics and the technical feasibility of remote sensing detection. Then, a detection model coupling the optical and thermal response characteristics is proposed to break through the bottleneck of oil-spill detection across time and space in complex scenes, which can enhance the diversity of samples by exploiting the 3-D spatial-spectral feature (SSF) generation adversarial expansion module constrained by the sun glint intensity index (SGII). Meanwhile, based on the thermal infrared super-resolution enhancement module guided by texture features, the thermal infrared spatial geometric features of oil spills are improved, and the model’s ability to resist cloud and fog interference is enhanced. Besides, under the guidance of the optical feature guidance (OFG) module, the SSF information can be extracted based on the SSF convolution deep belief network (SSF-CDBN) to realize the oil-spill remote sensing migration detection both temporally and spatially. The results show that the model can accurately detect oil spills and emulsion type, the overall accuracy (OA) and Kappa are larger than 84.46% and 0.758 in ideal scenarios, 80.11% and 0.745 in complex scenarios, which is expected to provide new technical support for oil-spill accidents.

Crude oil and its emulsions seriously threaten marine ecological environment and human health. Thermal infrared remote sensing technology is an important means of optical remote sensing marine oil-spill monitoring. In this paper, based on the UAV thermal infrared radiometer, a 24-hour continuous oil-spill brightness temperature image acquisition experiment was carried out, and oil-spill BTD polar coordinate thermal model was constructed to analyze the thermal infrared response characteristics of crude oil and emulsified oil. The results show that the thermal infrared response characteristics of oil-in-water (OW) emulsified oil and seawater are difficult to distinguish, which of crude oil and water-in-oil (WO) emulsion are similar. During 11:00~14:00, there is strong thermal infrared separability between oil and seawater, and the BTD values of crude oil and WO emulsified oil are both greater than 15. During 10:30~15:00, there is a strong positive correlation between BTD and oil film thickness, R2is greater than 0.9, so it is determined that the optimal time window for oil-spill thermal infrared monitoring is 11:00~14:00. The thermal infrared remote sensing monitoring based on oil-spill BTD model and optimal time window is expected to provide new technical and method support for marine oil-spill emergencies.

Oil film thickness (OFT) is an important indicator for estimating the amount of oil spill, and accurately quantifying the OFT is of great significance for loss assessment. In this paper, hyperspectral images (HSIs) of different OFTs (0.01-3.04 mm) through a ground experiment were obtained, and the spectral characteristics were analyzed. To address the issue of poor spectral separability for different OFTs, the 1DConvolutional Neural Network_Gate Recurrent Unit (1DCNN_GRU) model was developed for the quantitative inversion of OFT. It was validated through experiments on airborne Cubert-S185 HSI. The experimental results indicated that: (1) The proposed 1DCNN_GRU model effectively addressed the issue of reduced quantitative inversion accuracy resulting from poor spectral separability. The inversion results of it outperformed those of the SVR, CNN, and GRU models. Moreover, the optimal time for hyperspectral sensor to monitor OFT was at noon. (2) The proposed model using airborne hyperspectral data exhibited excellent inversion performance for OFT greater than 0.07 mm, especially with the best performance in 0.60-0.90mm. (3) The accuracy of HSI based OFT inversion assisted by brightness temperature (BT) data was superior to that of OFT inversion using single-source data. In particular, the proposed model had advantages in the feature level and decision level inversion of OFT in the ranges of 0.01-0.30mm and 1.00-3.04mm, respectively. This research provides technical support for the detection of OFT.

Water clarity (ZSD) is a critical water quality parameter that requires remote sensing mapping. Although great progress has been made inZSDretrieval over clear waters during past decades, challenges remain over turbid waters. To address this issue, a new model was proposed to retrieveZSDin clear oceanic to extremely turbid coastal waters, by improving theZSDretrieval in turbid waters. Firstly, waters were optically classified into three classes (clear, moderately turbid and extremely turbid waters) with band ratio of remote-sensing reflectance (Rrs(λ))f=Rrs(670)/Rrs(490). Secondly, class-specific algorithms were adopted to retrieve the spectral diffuse attenuation coefficientKd(λ) fromRrs(λ). Finally,ZSDwas semi-analytically estimated from minimumKd(λ) in the visible domain. Data from oceanic and coastal waters (N=2260) were used for the model parameterization, test and validation. To demonstrate the model applicability to major satellite sensors, 1299 images from six spaceborne sensors were matched up with independentin situ ZSD(N=1464,ZSD=0.2-51 m) from global oceans. The results indicate that the new model has a good performance with mean absolute percentage error (MAPE) and Root Mean Square Difference (RMSD) of 21%-26% and 0.3-2.8 m. Even over extremely turbid waters, the model still performs robustly (MAPE=22%-25%) and significantly better than the existing ones. Finally, the model application indicates that theZSDderived from six sensors show good agreement in both spatial distribution and temporal consistency. The model shows the potential to construct high-accuracyZSDrecords from multiple sensors for global oceans and can support sustainable management of marine ecological environment.

Marine oil-spill accidents seriously threaten both the marine ecological environment and human health. It is important to accurately identify the type of oil spills and detect the thickness of oil films on the sea surface to obtain the amount of oil spill for on-site emergency responses. Remote sensing is an important method for marine oil-spill detection and identification. In this study, thermal infrared remote sensing images of oil spills were obtained using thermal infrared imaging camera and UA V, and a marine oil-spill thermal infrared detection SVC model was proposed to conduct oil-spill detection research. The results of the land-based experiment show that there is a strong correlation between the thick oil film with different thicknesses and the surface temperature, and the R2 is larger than 0.92. The results of the UA V detection experiment show that the OA is larger than 76.84%, and the Kappa coefficient is larger than 0.740, which show the potential of UA V thermal infrared remote sensing for oil-spill detection.

Aiming at the needs and difficulties of shallow bathymetry in sea areas lacking in-situ depth information, taking Dongdao Island of Xisha Islands as an example, the active-passive fusion bathymetry using ICESat-2 photon counting lidar and WorldView-2 optical remote sensing image based on GRU deep learning model is studied. First, tidal corrections are carried out separately for multi-track ICESat-2 data, and the depth data are obtained through signal point extraction and refraction correction. Then, taking the depth data as control points, the GRU deep learning model is applied to carry out water depth inversion in the study area. The experiments show that the MAE and MRE of the bathymetric inversion for the whole area are 1.1m and 17.3% respectively. Furthermore, the MAE and MRE are 1.3m and 19.3% in a local area without Lidar data. Active-passive fusion bathymetry based on GRU model can provide a new means for shallow water depth detection.

Recently, marine oil spill incidents occur frequently, causing serious pollution, which has seriously endangered marine ecological environment security. The type of oil spill pollution is related to the formulation of punishment and cleaning scheme, which is an important basis for the disposal of oil spill pollution. Hyperspectral remote sensing is an effective means to monitor marine oil spills. Different types of light oils are difficult to identify effectively, which can not meet the needs of accurate monitoring applications. In this paper, the outdoor oil spill experiment is implemented. The data of five typical oil products are obtained by unmanned airborne hyperspectral imager, and the feature extraction and analysis are carried out. The RPnet deep learning recognition model of oil types under multi feature fusion is constructed to realize the effective identification of different oil spill types. It can provide important technical support for offshore oil spill monitoring of relevant business departments.

GF-1 multispectral wide field of view (WFV) images, with a spatial resolution of 16 m, have been widely used in earth monitoring. However, the spatial details provided by WFV images are not sufficient for many applications. Thus, this letter proposes a novel WFV image super-resolution (SR) algorithm called GFRCAN based on a very deep residual coordinate attention network. To form a very deep network, the residual-in-residual (RIR) structure consisting of several residual groups (RG) with long skip connections is used. Meanwhile, the residual coordinate attention block (RCOAB) and adaptive multi-scale spatial attention module (AMSA) are incorporated to focus on the high-frequency information and multi-scale features adaptive weighted fusion. Besides, the spectral and spatial details of SR images are improved by incorporating peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM) into the loss function. Both subjective and objective evaluation results show that the proposed model outperforms the state-of-the-art methods.

Ocean oil spill is an emergency with great harm. Optical remote sensing is an important means to monitor oil spill on the sea surface. Due to the influence of cloud and weather and the limitation of satellite revisit period, only limited sample data can be obtained. In the case of limited samples, the ability of learning sample features using a single supervised classifier is limited, which can not meet the needs of accurately monitor oil spill. This paper takes GF-1 WFV oil spill image as data source, and uses four classical supervised classification algorithms to extract oil spill information. From the point of view of target recognition information fusion, the advantages of multiple supervised classification algorithms are integrated. Decision fusion algorithm is used to realize multi-source oil spill information fusion, so as to improve the detection accuracy of remote sensing oil spill.

The phenomenon of spectral aliasing exists for coastal wetland object types, which leads to class mixing. This letter proposes a multiobject convolutional neural network (CNN) decision fusion classification method for hyperspectral images of coastal wetlands. This method adopts decision fusion based on fuzzy membership rules applied to single-object CNN classification to obtain higher classification accuracy. Experimental results demonstrate the effectiveness of the proposed method for the six object types, including water, tidal flat, reed, and other vegetation types. The overall accuracy of the decision fusion classification method based on fuzzy membership is 82.11%, which is 3.33% and 6.24% higher than those of single-object feature band CNN and support vector machine methods. The classification method based on multiobject CNN decision fusion inherits the characteristics of single-object feature bands of the CNN, making it a practical approach to image classification under the challenging conditions in which class mixing occurs.