Spaceborne synthetic aperture radar (SAR) imagery reveals that the leakage of heavy fuel from a sunken shipwreck on the Bulgarian Black Sea coast had been ongoing for more than ten years, before it was noticed in mid-2018. Our analyses of long-term spaceborne SAR data indicate that the heavy fuel formed an oil spill identifiable on SAR imagery only under favorable environmental conditions, when the temperature of the sea water was high enough, after periods of strong winds, heavy sea state, and strong sea bottom currents, and when the wind speed during SAR image acquisition was low enough to allow the formation of a coherent spill.

Intertidal flats are vulnerable areas that are difficult to access and that are home to many species. We are using synthetic aperture radar (SAR) data for the monitoring of exposed intertidal flats to assess their economic use through intensive aquacultures, and to identify areas of strong morphodynamics. Polarimetric decomposition of Gaofen-3 SAR data of various spots on the Chinese coast revealed that aquacultural rafts show different backscattering mechanisms than the surrounding mud flats. Moreover, Sentinel-1 SAR-C data are used to generate topographic maps of extended intertidal flats on the German North Sea coast. Those maps, derived for different time periods, will be used to assess morphodynamic changes.

Dual-co-polarization Synthetic Aperture Radar (SAR) imagery also has large potentials in detecting sediments and habitats on exposed intertidal flats. We expand on our previous works of FCD-RF algorithm, that is, Freeman-Durden and Cloude-Pottier decompositions of polarimetric SAR, along with a Double-Bounce Eigenvalue Relative Difference (DERD) parameter, and a Random Forest (RF) classifier. By adding indicators derived from the (polarimetric) Kennaugh elements, that is, FCDK-RF, we found its good performance in distinguishing sand flats, mud flats, and bivalves. We tested our results using Radarsat-2 (C-band) dual-co-polarization$(\text{HH}+\text{VV})$SAR data acquired along the German Wadden Sea. Results show that FCDK-RF fulfills the requirements in intertidal flats mapping projects especially for sediments and haibitats.

We propose a SAR and Optical Fusion Network based on the UNet framework (SOF-UNet) for multi-modal land cover classification. The two-stream SOF-UNet consists of three parts: two encoders to extract features, a sharing decoder to upsample the feature maps and specially designed skip connections to fuse multi-modal features. The qualitative and quantitative experimental results show that SOF-UNet has a promising capability to identify different land cover classes and can retain fine details in the prediction maps. Symmetric Cross Entropy (SCE) loss is also verified useful in this framework.

Marine surface films, be they of biogenic or anthropogenic origin, dampen the small-scale waves that are responsible for the radar backscattering, hence they can be seen on synthetic aperture radar (SAR) imagery as areas of reduced image brightness. Examples of different kinds of marine surface films are presented, that were observed on spaceborne SAR imagery since the launch of SEASAT in 1978. In addition, some results from SAR data analyses are used to demonstrate the key aspects of the visibility of marine surface films on spaceborne SAR imagery.

We use 14 years of spaceborne SAR imagery from Sentinel-1A/B, ALOS-1/2 PALSAR-1/2, and Envisat ASAR to show that heavy fuel was leaking out of a ship wreck off the Bulgarian Black Sea coast. The oil spills were detected only in the summer months, when the sea bottom temperature exceeded 10 °C, indicating that the viscosity of the heavy fuel is too high at lower temperatures. Moreover, periods of high winds, hence of high sea state, were preceding each oil spill detecting, which indicates that mechanical action on the wreck's hull was needed to initiate further leakage.

The objective of this research is to realize automatic land cover classification from synthetic aperture radar (SAR) and multispectral remote sensing imagery. We develop a SAR-optical fusion network (SOFNet) with the symmetric cross entropy (SCE) loss to utilize both the SAR and optical information in a novel deep neural network. The proposed framework has been trained on the public SEN12MS dataset and tested on the 2020 IEEE-GRSS Data Fusion Contest (DFC2020) dataset. Experimental results show that our approach takes full advantage of multimodal information and outperforms the state-of-the-art convolutional architectures.

This paper demonstrates how longterm spaceborne SAR imagery of coastal zones can be used to study morphodynamical changes in intertidal areas. Depending on local environmental conditions exposed intertidal flats may show up as dark or bright patches on SAR imagery, thereby allowing the identification, and the tracking, of changes in the sediment distribution that are related to anthropogenic impacts.

We used about 2000 Sentinel-1 SAR-C images of the Western Java Sea for statistical analyses of the marine oil pollution in that area. Our results demonstrate that most pollution was found along the main ship traffic routes, and further hot spots of oil pollution were found in offshore oil fields. Although the mean wind speed in the area of interest is generally low enough to support the visibility of oil spills on SAR imagery, we found a reverse correlation with the number of oil spills per SAR image. About 50% of the oil spills covered a sea surface area up to 0.5 km2, and another 50% had an axes ratio up to 10, i.e., they were elliptical or quasi-circular. Comparing results obtained by different operators we found that the absolute numbers of oil spills differed, but the qualitative conclusions, regarding the location of pollution hot spots, are the same.

The objective of this research is to detect ocean eddies automatically on Synthetic Aperture Radar (SAR) images. We develop a new approach using Mask Region-based Convolutional Neural Networks (Mask R-CNN) and edge enhancement. First, we use Canny edge detector to extract a wide range of edges in SAR images. Then we put both the edge detection results and the corresponding original images into a Mask R-CNN based model for learning, thereby strengthening edge information. The proposed framework has been trained on a sample dataset of Sentinel-1A SAR-C imagery of the Western Mediterranean Sea. Experimental results revealed that the proposed method improved the performance by 2.3% on the MS COCO metrics compared to the method without edge enhancement.

We show that heavy fuel leaking out of an old ship wreck 10 km off the Bulgarian Black Sea coast can be identified on spaceborne Synthetic Aperture Radar (SAR) imagery. Sentinel-1A SAR-C and ALOS-2 PALSAR-2 imagery is used to demonstrate that the leakage started more than three years ago, in summer 2015, hence it has been going on for a much longer time than stated by local experts. The resulting oil spills can have lengths exceeding 9 km, and they may be (partly) displaced due to strong subsurface currents.

We inferred spatial densities of mesoscale and submesoscale eddies in the Western Mediterranean from Sentinel-1 SAR-C imagery. Highest normalized eddy densities, i.e. average numbers of eddy manifestations per SAR image acquired under favorable wind conditions, were found in autumn in several parts of the Western Mediterranean, when the biological enrichment of surface waters is high enough, the regional mean wind field is variable and when regional mean SST gradients are strong enough. Our results also imply that the observation period of 3 years is not sufficiently long for definite conclusions.

Quad-polarization Synthetic Aperture Radar (SAR) imagery can be well used to identify bivalve beds on dry-fallen intertidal flats and to investigate radar backscattering mechanisms on exposed sediments. In this study, a general classification approach to use spaceborne SAR data to distinguish sediment types, especially mud and sand, and to detect mussel beds, is presented. We propose a new feature set including Double-bounce Eigenvalue Relative Difference (DERD) and the Kennaugh elements, and combine those indicators with the unsupervised Wishart classifier. The main classification results of sediments and habitats agree with data from regular monitoring campaigns, which are frequently performed to fulfill the requirements of coastal environmental monitoring.

We inferred spatial densities of submesoscale eddies in the Western Mediterranean from EnviSat ASAR and Sentinel-I SAR-C imagery. The derived normalized eddy densities, i.e. the average numbers of eddy manifestations per SAR image acquired under favorable wind conditions, for the two datasets show some similarities, but they also demonstrate that the observation periods of 2-3 years are not sufficiently long for definite conclusions.

We demonstrate that dual-co-polarization Synthetic Aperture Radar (SAR) imagery can be used to derive new indicators for bivalve beds on dry-fallen intertidal flats and to investigate radar backscattering mechanisms on exposed sediments and seagrass meadows. The Kennaugh element framework forms the basis for our studies, and particular Kennaugh elements are used to generate maps indicating the location of bivalve beds and to demonstrate the seasonal change of the radar backscattering from exposed intertidal during low tide.

We analyzed more than 5000 synthetic aperture radar (SAR) images from ERS 1/2, ENVISAT, ALOS-1, and Sentinel-1A with respect to marine oil pollution in two regions of interest (ROIs) in Indonesian waters. Both normalized spill number and normalized mean polluted area indicate that the marine oil pollution in the ROIs generally is of different origin: while ship traffic appears to be the main source in the Java Sea, oil production industry causes highest pollution rates in the Strait of Makassar. Largest oil spill numbers in the Java Sea were found during the transition from north-west to south-east monsoon, and vice versa, when the overall wind and current patterns change, thereby making oil pollution detection with SAR easier. We further demonstrate that high-resolution numerical forward and backward tracer experiments, in combination with (visual) SAR image analyses help in better understanding the observed seasonality.

The most common polarimetric decompositions can only be computed for fully polarimetric data at the cost of reduced spatial resolution and areal coverage. In this paper, we analyze the potential of dual-pol SAR imagery for the monitoring of bivalve beds on intertidal flats. The normalized Kennaugh elements, which can be calculated from dual-pol data of any wavelength, are applied to deduce polarimetric information from HH and VV polarized TerraSAR-X imagery. The real (K3) and imaginary (K7) parts of the inter-channel correlations are demonstrated to be indicators for bivalve (oyster and mussel) beds in a test site in the German Wadden Sea. Our results show that K3 and K7 can be used to detect bivalve beds with high accuracy.

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.

We analyzed a large amount of high-resolution Synthetic Aperture Radar (SAR) data of exposed intertidal flats on the German North Sea coast with respect to the imaging of sediments, macrophytes, and mussels. TerraSAR-X and Radarsat-2 images of four test areas acquired from 2008 to 2013 form the basis for the present investigation and are used to demonstrate that pairs of SAR images, if combined through basic algebraic operations, can already provide indicators for bivalve (oyster and mussel) beds. Our results show evidence that single-acquisition, dual-polarization SAR imagery can be used in this respect. The polarization coefficient (i.e., the normalized difference polarization ratio) can be used to infer indicators for oyster and blue-mussel beds.

In the frame of the German-Indonesian pilot study IndoNACE (Indonesian Seas Numerical Assessment of the Coastal Environment), a wealth of SAR data of two dedicated regions in Indonesian waters are being analysed with respect to the imaging of marine oil pollution. Numerical tracer studies using a regional 3-d numerical model are used to aid those analyses and to help understanding the observed seasonal variations in marine oil pollution. Our first results are based on 130 ENVISAT ASAR images of each of the two regions of interest, the `Western Java Sea' and the `Makassar Strait' and indicate that most pollution was found in areas of high ship traffic and of intense oil production.

In the present study we analyze manifestations of submeso- and mesoscale eddies in the Western Mediterranean Sea found in Envisat Advanced Synthetic Aperture Radar (ASAR) imagery of 2010-2011. Location of mesoscale eddies found in the imagery outlined the main stream of the basin. Special attention is being paid to the relation between the numbers of cyclonic and anticyclonic eddies at different spatial scales. It is shown that eddies less than 60 km in diameter are dominantly cyclonic; at some intermediate scales (about 70 km) there is a balance between the numbers of cyclonic and anticyclonic eddies and at largest scales (>100 km) eddies are predominantly anticyclonic.

After major storm surges in the 14thand 17thcenturies, vast areas on the German North Sea coast were lost to the sea. However, remnants of former settlements and of historical landuse, which were buried by sediments for several centuries, can still be found in the German Wadden Sea, if the surface layer is driven away under the permanent action of wind, currents, and waves. We show that high-resolution SAR imagery with pixel sizes well below 1 m2can be used to complement archeological surveys and that TerraSAR/ TanDEM-X images clearly show remnants of farmhouse foundations and of former systems of ditches, dating back to the 14thcentury and to the 16th/17thcentury. In particular, the new high-resolution TerraSAR-X acquisition mode (`staring spotlight') allows for the detection of various kinds of residuals of historical landuse, some of which have been unknown so far.

We discuss the spatio-temporal distribution of sub-mesoscale eddies seen in synthetic aperture radar (SAR) imagery of the Baltic Sea. Analyzing about 1250 Envisat Advanced SAR (ASAR) images acquired between 2009 and 2011 almost 7000 sub-mesoscale eddies were discovered. Since the visibility of vortical structures in SAR imagery significantly depends on the near-surface wind speed, wind data from a numerical model of the Baltic Sea were additionally used to improve our eddy statistics. Within the method proposed herein only those parts of SAR images are considered for the calculation of eddy statistics, which were acquired when the wind speed conditions were favorable for eddy manifestations. As a result, and despite the fact that eddies were generally observed all over the Baltic Sea, we show that the south-western part of the Baltic Sea seems to have especially high sub-mesoscale eddy activity.

High-resolution Synthetic Aperture radar (SAR) images of coastal and off-coastal areas provide a nearly weather independent monitoring of small-scale oceanographic features. In this paper, we describe the advances, which have been achieved within the DTeddie project. Therefore, we present the application of formerly developed algorithms to high-resolution SAR data to estimate the sea surface current at sub-mesoscale. Inside the derived current sea surface current fields, vortical structures, e.g. eddies, will be detected by means of local current anomalies. Another approach uses the alignment of anchored ships as an indicator of the sea surface current. To infer dependent information for each ship, we present an automatic registration method between coastal photographs and the SAR images. Both approaches have proven to yield valuable results with respect to the detection of vortical current flows at sub-mesoscale.

The possibility of detecting breaking long surface gravity waves in S-band radar backscatter measurements is being investigated using multi-polarization scatterometer measurements. To detect breaking waves the change in the spectral shape of the radar Doppler spectra was analyzed with respect to possible indicators for wave breaking events. They are identified here using the ratio of the second and fourth spectral moments after normalization by the squared radar frequency. A comparison with simultaneous estimates of the fractional area of whitecaps suggests that a normalized spectral moment ratio might be used as one indicator for events of breaking surface gravity waves.

High-resolution Synthetic Aperture Radar (SAR) data of dry-fallen intertidal flats in the German Wadden Sea have been analyzed with respect to the imaging of sediments, macrophytes, and mussels. A large number of high-resolution TerraSAR-X and Radarsat-2 images of five test areas along the German North Sea coast were acquired in 2012 and 2013 and form the basis for the present investigation. Depending on the type of sediment, but also on the water level and on environmental conditions (wind speed) exposed sediments may show up on SAR imagery as areas of enhanced, or reduced, radar backscattering. The (multi-temporal) analysis of series of such images allows for the detection of mussel beds, and our results show evidence that also single-acquisition, multi-polarization SAR imagery can be used for that purpose.

In [6] we presented an approach to measure sea surface currents from satellite images by computing the displacement vectors of distinctive image features or measuring the optical flow between successive images of the same scene. These algorithms are based on the assumption that the image flow observed is caused by the motion of surface films with the currents to be measured. Optical flow due to other image features, e.g. ship wakes, does not correspond to the motion field and causes systematic errors. The decision whether a derived measurement is valid or not by domain experts is very time-consuming. We present first steps towards an automatic knowledge-based approach that uses description logic to validate the measurements of sea surface currents. The terminological knowledge is based on the expert's knowledge of the domain in conjunction with a geographical database containing factual knowledge about the scene. The combination of different sources of knowledge makes it possible to infer about the validity of sea surface current measurements from image data by reasoning based on the local context of the image features.

Sequential multi-sensor satellite images are used for the computation of mesoscale surface currents in the Northern and Southern Baltic Proper. Different marine surface films and accumulated algae at the water surface are imaged by the sensors working in the optical, infrared, and microwave part of the electromagnetic spectrum and can thus be used as tracers for the local motion of the sea surface. Due to the generality of the problem, there are different algorithms available for motion estimation using object tracking in images. To apply these to the field of sea surface current estimation, we need to take the smoothness assumptions into account that the algorithms were built up on. We present the influence of different spatial constraints in the algorithms for the derivation of mesoscale sea surface currents using multi-sensor / multi-channel satellite images by means of a quantitative comparison with model results provided by the German Federal and Maritime Agency (BSH).

A helicopter-based Doppler scatterometer (Multi$^{3}$Scat) is described. It allows simultaneous measurements of the surface radar backscatter at five different frequencies at co- and cross-polarization at incidence angles of 20$^{\circ}$–65$^{\circ}$from an altitude of 30–300 m. Video and infrared (IR) cameras simultaneously sense the surface in the scatterometers' footprint. The Multi$^{3}$Scat is calibrated using measurements carried out over corner reflectors. The stability of the Multi$^{3}$Scat's signal is found to be, on average, better than 0.5 dB. Typical signal-to-noise-ratio values for sigma-0 range between 10 and 20 dB for cross-polarization and between 15 and 25 dB for copolarization over snow and ice surfaces. The potential of the Multi$^{3}$Scat to acquire multifrequency multipolarization radar backscatter data and coincident video and IR temperature observations at different incidence angles over remote terrain such as the Arctic Ocean or the Alps is demonstrated.

Microwave modulation by swell waves and its relation to microbreaking waves were investigated in an ocean experiment. Simultaneous collocated microwave and infrared (IR) measurements of wind waves and swell on the ocean were made. The normalized radar cross section sigma0and the skin temperatureTskinwere both modulated by the swell, but with differing phases. In general, sigma0maxima occurred on the front face, whereasTskinmaxima occurred on the rear face of the swell. Infrared imagery has shown that swell-induced microbreaking occurred at or near the swell crest and that the resulting warm wakes occurred on the rear face of the wave. When tilt and range modulations are taken into account, the location of microbreaking also accounts for the maximum of sigma0occurring on the front face of the swell. Thus, microbreaking waves generated near the crest of low-amplitude swell can produce microwave and IR signatures with the observed phase. The relationship between microwave and IR signals was further emphasized by comparing microwave Doppler spectra with simultaneous IR and visible images of the sea surface from the same location. When small and microscale breaking waves were present, Doppler spectra exhibited characteristics that are similar to those from whitecaps, having peaks with large Doppler offsets and polarization ratios near unity. When no microbreakers were present, Doppler offsets and polarization ratios were much smaller in accordance with a composite surface scattering theory.

Coastal surface currents have been computed for years from sequential infrared and more recently ocean color imagery using the Maximum Cross Correlation (MCC) technique. Preliminary results suggest that this MCC method may be applied to sequential Synthetic Aperture Radar (SAR) imagery yielding surface currents with a much higher spatial resolution which are independent of the presence of cloud cover which makes it impossible to use infrared or ocean color imagery. A requirement for the application of the MCC to SAR imagery is the presence of surface slicks, which are often related to ocean color patterns. Test applications are made to ENVISAT ASAR images.

The different morphology of palmitic acid metliyl ester (PME) slicks spread from n-hexane or ethanol was studied by 'Brewster Angle Microscopy' [BAM]. In the latter case, already at large areas per molecule a foam-like structure is being formed on the water surface, i.e., a two-dimensional network which appears to be comparable with the morphological structure of biogenic sea slicks. In line with this assumption, both the magnitudes of the radar backscatter damping ratios and the characteristics of the damping ratio/wavenumber curves were comparable for the PME slick spread from ethanol and for the biogenic slicks, while in the presence of the PME slick spread from n-hexane lower damping ratios were determined. Furthermore, the relaxation of alkanoic acid esters, which are often being found in biogenic sea slicks, was investigated by 'Infrared Reflection-Absorption Spectroscopy' [IRR4S]. It turned out that the ester group is continuously hydrated and dehydrated during compression and dilation on an undulating water wave field. If can be safely assumed that the strong water wave damping induced by these chemical compounds is centrally related to this phenomenon.

In order to gain further insight into the radar backscattering from rain-and wind-roughened water surfaces, we performed field experiments at the mouth of the river Elbe in the German Bight. The backscattered radar power from the water surface was measured at different wind speeds, wind directions, and rain rates, as well as at different radar frequencies, polarizations, and incidence angles. Our results show that the dominant scattering mechanism at VV-polarization is Bragg scattering which shows a strong dependence on the Bragg wavelength. The enhancement and reduction of the surface roughness by splash products and sub-surface phenomena, respectively, result in a transition wavenumber between enhancement of the radar backscattering (due to additional surface roughness) and its reduction (due to wave damping) of about 100 rad/m. We hypothesize that this transition wavenumber depends on the drop-size distribution of the rain. Taking into consideration the different dependencies on rain rate of the radar backscatter at different frequencies and polarizations, we suggest a method to estimate rain rates by calculating the ratio of the radar cross-sections at L band, VV-polarization and at C band, HV-polarization. Provided an availability of synthetic aperture radar (SAR) data at the respective frequency/polarization combinations, this method will allow for investigating the nature of small-scale (convective) rain events over the ocean by analyzing multi-frequency/multi-polarization SAR images.

We used multi-sensor satellite images to compute small-scale surface currents in the Northern Baltic Sea. The images were acquired by the Thematic Mapper (TM) aboard Landsat, by the Synthetic Aperture Radar (SAR) aboard ERS-2, and by the Wide-Field Scanner (WiFS) aboard IRS-1C within two hours on 15 July, 1997, over the Northern Baltic Proper. Most of them show manifestations of an ongoing algae bloom, which are driven by the local currents. Pairs of images acquired at either different electromagnetic frequencies (TM and SAR) or at the same electromagnetic bands (TM and WiFS) are used for applying cross-correlation or partial differentiation techniques, respectively. Our results are insofar promising as the computed two-dimensional surface current fields complement existing data from numerical models. However, limitations of the proposed method are due to the strong dependence of the visibility of algae bloom manifestations on local weather conditions and to the low availability of satellite data.

We have carried out wind-wave tank measurements to investigate the small-scale roughness of the water surface at the edges of monomolecular surface films. Small amounts of oleyl alcohol, triolein, and palmitic acid methyl ester were carefully deployed on the water surface at wind speeds between 3 m/s and 9 m/s, thus allowing for wave measurements at either end (i.e., downwind and upwind) of the surface films. Data were acquired using a 10 GHz (X-band) scatterometer, and a two-dimensional laser slope gauge, and a video camera. We found that the effect of slick-edge asymmetry (i.e., a sharp downwind edge and a diffuse upwind edge) depends on both the deployed substance and the wind speed. Due to different states of film compression at the downwind or upwind edges of the surface film, the damping capability of the slick material can vary in these areas. This in turn may cause a varying spectral density of small-scale waves, which are bound to the dominant wind waves, while the slick is drifting by. This effect is shown and interpreted qualitatively using the video data, and quantitatively using the scatterometer and laser gauge data.

W-band backscatter experiments have been performed at the wind-wave tank of the University of Hamburg. The aim of the experiments is to study the backscatter mechanisms under shallow incidence angles and to compare the results with previous ones obtained with an X-band scatterometer. The measurements were performed at three grazing angles, 7.5/spl deg/, 10/spl deg/, and 20/spl deg/, and at VV-, HH-, and VH-polarization. The wind speed ranged from 2 to 10 ms/sup -1/ and was increased in steps of 1 ms/sup -1/. The analysis of the acquired radar Doppler spectra shows that Bragg scattering from both bound and freely propagating Bragg waves is the dominant backscattering mechanism at all deployed wind speeds. In particular, at low wind speeds (up to 4 ms/sup -1/), when the overall backscatter is small, bound waves axe the dominating scatterers, whereas at higher wind speeds (5 ms/sup -1/ and above) the acquired signal is mainly caused by freely propagating Bragg waves. This is in qualitative agreement with similar findings, which have been made earlier with an X-band scatterometer working at moderate incidence angles.