No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
A physically consistent stochastic model to observe oil spills and strong scatterers on SLC SAR images
Abstract
A speckle model to characterize low backscatter areas and areas with strong scatterers in marine SLC SAR images is presented. The model allows using high resolution speckled SAR images instead of dealing with multi-look SAR images where, at the expense of a poorer spatial resolution, the speckle is mitigated. The new approach is based on the use of the three parameters of the generalized K probability density function. This speckle model embodies the Rayleigh, the Rice and the K-distribution scattering scenes, which are descriptors of scenes dominated by Bragg scattering, scenes in which a dominant scatter is present and scenes with a non-Gaussian signal statistic, respectively. A large data-set of ERS 1/2 SLC SAR images, provided by the ESA under the Project C1P-2769, is employed. Results show the effectiveness of the approach.
... It is a three parameter distribution in which a further parameter, , is present in addition to the two previous ones. The real and nonnegative parameter takes into account the departure from the uniform distribution of the phase, e.g., a phenomenon which occurs when a dominant scatter is present [15], [20]. When , then and the strong scattering regime is achieved. ...
... When , then and the strong scattering regime is achieved. The GK model has been shown to embody very different marine scattering scenarios, each one being characterized by a proper combination of GK parameters [15], [20]. From an electromagnetic viewpoint, the mean intensity of the total field scattered off the sea surface with and without metallic objects, , is associated with the sum of a coherent and incoherent component, each properly weighted by the Rice factor [21] (3) ...
An electromagnetic model, based on the generalized -K (GK) speckle distribution, has been developed to read the scattering features associated with metallic objects at sea in full resolution single-look complex (SLC) synthetic aperture radar (SAR) data. The sensitivity of the GK parameters is investigated in both copolarized and cross-polarized SAR data. It is shown that a proper combination of GK parameters exhibits a completely different behavior with respect to sea surface with and without metallic objects, when measured over cross-polarized SAR data. As a matter of fact, a simple and very effective approach to observe metallic objects in full resolution SLC cross-polarized SAR data, is proposed. Experiments undertaken over a large data set consisting of RADARSAT-2 SLC fine quad polarization SAR data, confirm model predictions, and show that the proposed approach is both physically-based and operationally effective, since it is able to provide a logical true and false output.
... It is a three parameter distribution in which a further parameter, , is present in addition to the two previous ones. The real and nonnegative parameter takes into account the departure from the uniform distribution of the phase, e.g., a phenomenon which occurs when a dominant scatter is present [15], [20]. When , then and the strong scattering regime is achieved. ...
... When , then and the strong scattering regime is achieved. The GK model has been shown to embody very different marine scattering scenarios, each one being characterized by a proper combination of GK parameters [15], [20] . From an electromagnetic viewpoint, the mean intensity of the total eld scattered off the sea surface with and without metallic objects, , is associated with the sum of a coherent and incoherent component , each properly weighted by the Rice factor [21] (3) where is the Rice factor, i.e., the coherent to incoherent mean intensity eld ratio [21], [22] (4) A metallic object over the sea surface calls for a strong electromagnetic return which is, at once, characterized by a large coherent component [8]. ...
An electromagnetic model, based on the generalized-K (GK) speckle distribution, has been developed to read the scattering features associated with metallic objects at sea in full resolution single-look complex (SLC) synthetic aperture radar (SAR) data. The sensitivity of the GK parameters is investigated in both copolarized and cross-polarized SAR data. It is shown that a proper combination of GK parameters exhibits a completely different behavior with respect to sea surface with and without metallic objects, when measured over cross-polarized SAR data. As a matter of fact, a simple and very effective approach to observe metallic objects in full resolution SLC cross-polarized SAR data, is proposed. Experiments undertaken over a large data set consisting of RADARSAT-2 SLC fine quad polarization SAR data, confirm model predictions, and show that the proposed approach is both physically-based and operationally effective, since it is able to provide a logical true and false output.
... SeaWinds observations (Mercier and Girard-Ardhuin 2005;Quintero-Marmol et al. 2003;Migliaccio et al. 2007;Shao et al. 2008;Mercier and Ardhuin 2006b), underwater gliders for in-situ ocean measurements, and Cross Calibrated Multi-Platform (CCMP) wind data (Tian, Huang, and Hongga 2017). Moreover, in some of the papers, these conditions were estimated and retrieved from SAR images using the CMOD4/5 model (CMOD is a C-band geophysical model that provides an empirical relation between the radar backscatter sensed from the roughened sea surface and wind speed) (Najoui et al. 2018;Vijayakumar and Rukmini 2016;Mera et al. 2017;Hersbach, Stoffelen, and de Haan 2007;Garcia-Pineda et al. 2013;Kim et al. 2015). ...
Oil spills are one of the most hazardous disasters with significant short-and long-term effects on fragile marine ecosystems. Synthetic Aperture Radar (SAR) has been considered an effective technology for mapping and monitoring oil spills in the marine environment, primarily thanks to its weather-, illumination-, and time-independent capabilities. To cope with serious oil spill threats, researchers have developed various analytical methodologies utilizing key advantages of SAR imagery to identify the occurrence of oil spills and discriminate lookalikes. Choosing the appropriate SAR specifications and investigating the effects of field conditions are challenging for oil spill monitoring and should be investigated further. This paper presents a comprehensive review study on maritime surveying and oil slick detection using SAR imagery through indexed research studies' compilation and analysis. To this end, a total of 230 peer-reviewed papers, published in various remote sensing (RS) journals and 78 conference papers in the International Society for Photogrammetry and Remote Sensing (ISPRS) archive and the IEEE International Geoscience and Remote Sensing Symposium (IGARSS) proceedings were reviewed. Our review study represents a meta-analysis investigation of these papers focusing on several features, including data, sensor type, imaging mode, microwave carrier frequency (e.g., L-, C-, and X-bands), polarization option (i.e., single-pol, dual-pol, full-pol, and compact-pol), incidence angle, and wind speed. Furthermore, it provides an overview of the RS techniques developed to deal with the oil spill detection task. This paper can be a guideline for two groups of audiences; those interested in oil spill monitoring who want to get an overview of the problem and how to address it, and those already working in the field who want to understand the scope of the work being accomplished. Consequently, the current paper contributes both to academic RS research and to practical applications. ARTICLE HISTORY
... Therefore, several algorithms for removal of speckle have been developed [170]. Since speckle is temporary, multi-look imaging is one way to decrease speckle by a large amount [171,172]. ...
Remote-sensing for oil spills is reviewed. The technical aspects of sensors are summarized and the benefits and limitations of each sensor are given. The use of visible techniques is ubiquitous; however, it gives only the same results as visible monitoring. Oil has no particular spectral features that would allow for identification among the many possible background interferences. Identification of specific oil types is not possible. Cameras are useful to provide documentation.
Infrared (IR) offers some potential as an oil spill sensor. In daytime oil absorbs light and remits a portion of this as thermal energy at temperatures 3 to 4 K above ambient. IR cameras are economical, however they suffer from problems such as the inability to discriminate oil from interferences on beaches, among weeds, debris or sediment, and under certain lighting conditions. Furthermore, water-in-oil emulsions may not be detected in the infrared, depending on light and sea conditions.
The laser fluorosensor is a useful instrument because of its unique capability to identify oil on backgrounds that include water, soil, weeds, ice and snow. It is the only sensor that can positively discriminate oil on most backgrounds. The laser fluorosensor also allows for positive identification and discrimination between oil types.
Radar detects oil on water because oil dampens water-surface capillary waves under low to moderate wave/wind conditions. Radar offers the only potential for large area searches, day/night and foul weather remote sensing. False targets can be as high as 95%. Satellite-borne radar sensors are now extensively used for mapping large spills or assisting in ship and platform discharge monitoring.
The only commercial equipment that measures slick thickness is passive microwave.
... Migliaccio and group studied the processing of SAR images from an aircraft-based sensor [135,137,138]. It was noted that the main obstacle to analysis was speckle in the images. ...
... Therefore, several algorithms for removal of speckle have been developed (Liu et al., 2013). Since speckle is temporary, multi-look imaging is one way to decrease speckle by a large amount (Migliaccio et al., 2007;Migliaccio, 2005). ...
Even though the design and electronics of sensors are becoming increasingly sophisticated and sensors are becoming much less expensive, the operational use of remote sensing equipment lags behind the development of the technology. The most common forms of oil spill surveillance and mapping is done with simple still or video photography, which provide little, if any, forensic data. Remote sensing from aircraft is still the most common form of oil spill tracking. Attempts to use satellite remote sensing for oil spills, although successful, are not necessarily as claimed and are generally limited to identifying features at sites of known oil spills. The laser fluorosensor is a most useful instrument to forensics because of its unique capability to positively identify oil against most backgrounds, including water, soil, weeds, ice, and snow. Radar offers the only potential for searching in large areas and carrying out remote sensing during foul weather conditions, but offers very poor positive detection characteristics and thus low forensic capability. The usefulness of the visible spectrum for oil detection is limited. It is, however, an economical way to document oil spills and provide baseline data on shorelines or relative positions.
... Therefore, several algorithms for removal of speckle have been developed [170]. Since speckle is temporary, multi-look imaging is one way to decrease speckle by a large amount [171,172]. ...
Remote-sensing for oil spills is reviewed. The technical aspects of
sensors are reviewed and the benefits and limitations of each sensor are
given. Oil spill response often requires that remote sensing is used to
detect and map the spill of interest. A wide variety of technologies had
been tried.
The Italian vessel Costa Concordia wrecked on January 13th 2012 offshore the Giglio Island (Tuscany, Italy), with the loss of 32 lives. Salvage operation of the vessel started immediately after the wreck. This operation was the largest and most expensive maritime salvage ever attempted on a wrecked ship and it ended in July 2014 when the Costa Concordia was removed from the Giglio Island, and dragged in the port of Genoa where it was dismantled. The refloating and removal phases of the Costa Concordia were monitored, in the period between 14th and 27th of July, exploiting SAR (Synthetic Aperture Radar) images acquired by the X-band COSMO-SkyMed satellite constellation in crisis mode. The main targets of the monitoring system were: (i) the detection of possible spill of pollutant material from the vessel and (ii) to exclude that oil slicks, illegally produced by other vessels, could be improperly linked to the naval convoy during its transit along the route between the Giglio Island and the port of Genoa. Results point out that the adopted monitoring system, through the use of the COSMO-SkyMed constellation, can be profitably employed to monitor emergency phases related to single ship or naval convoy over wide areas and with a suitable temporal coverage. Furthermore, the refloating and removal phases of the Costa Concordia were a success because no pollution was produced during the operations.
This chapter outlines several general reviews of oil spill remote sensing. Massive spills of oil and related petroleum products can have serious biological and economic impacts. Following a spill, there is a scrutiny both from public and media with demands that the location and extent of the oil spill be determined. This chapter discusses various sensors. Remote sensing is playing a major role in determining the extent of oil spill. With the help of this instrument, oil can be monitored on the open ocean around the clock. The demerit of this instrument is that it lags behind in technology. Another instrument known as 'optical sensor' helps in determining the extent of oil spill. Optical techniques, the most common means of remote sensing uses the same range of the visible spectrum detection. Of late, a new technology which is known as Global Positioning Systems (GPS) is directly used to map remote-sensing data onto base maps. This chapter reviews the future trends of remote sensing such as cameras and thermal IR cameras, laser technology. Also rapidly improving computer capabilities allows real-time processing. Combinations of airborne and satellite-borne sensor systems are recommended in order to respond effectively to major marine oil spills.
Real and synthetic aperture radars have been used in recent years to image ocean surface waves. Though wavelike patterns are often discernible on radar images, it is still not fully understood how they relate to the actual wave field. The present paper reviews and extends current models on the imaging mechanism. Linear transfer functions that relate the two-dimensional wave field to the real aperture ra- dar (SLAR) image are calculated by using the two-scale wave model. It is noted that a description of the imaging process by these transfer functions can only be adequate for low to moderate sea states. Possible other mechanisms that contribute to the visibility of waves by real aperture radar at higher sea states, such as Bragg scattering from spontaneously generated short waves at peaked crests or in wave breaking regions, and Rayleigh scattering from air bubbles entrained in the water and from water droplets thrown into the air by breaking waves, are discussed in a qualitative way. The i_m_agLn_g mecha.nism for synthetic aperture radars (SAR's) is strongly influenced by wave motions (i.e., by the orbital velocity and accelera- tion associated with the long waves). The phase velocity of the long waves does not enter into the imag- ing process. Focusing of ocean wave imagery is attributed to orbital acceleration effects. The orbital mo- tions lead to a degradation in resolution which causes image smear as well as a SAR inherent imaging mechanism called velocity bunching. The parameter range for which velocity bunching is a linear map- ping process is calculated. It is shown that linearity holds only for a relative small range of ocean wave parameters: The likelihood that the transfer function is linear increases as the direction of wave propaga- tion approaches the range direction, as the wavelength increases, and as the wave height decreases. Lin- earity is required for applying simple linear system theory for calculating the ocean wave spectrum from the gray level intensity spectrum of the image. Although, in general, the full ocean wave spectrum cannot be recovered from the SAR image by applying simple linear inversion techniques, it is concluded that for many cases in which the ocean wave spectrum is relatively narrow the dominant wavelength and direc- tion can still be retrieved from the image even when the mapping transfer function is nonlinear. Finally, we compare our theoretical models for the imaging mechanisms with existing SLAR and SAR imagery of ocean waves and conclude that our theoretical models are in agreement with experimental data. In particular, our theory predicts that swell traveling in flight (azimuthal) direction is not detectable by SLAR but is detectable by SAR.
A methodology for retrieving high-resolution ocean surface wind fields from satellite-borne synthetic aperture radar (SAR) data is introduced and validated. The algorithms developed are suited for ocean SAR data, which were acquired at the C band of either vertical (VV) or horizontal (HH) polarization in transmission and reception. Wind directions are extracted from wind-induced streaks that are visible in SAR images of the ocean at horizontal scales greater than 200 m. These wind streaks are very well aligned with the mean surface wind direction. To extract the orientation of these streaks, two algorithms are introduced, which are applied either in the spatial or spectral domain. Ocean surface wind speeds are derived from the normalized radar cross section (NRCS) and image geometry of the calibrated SAR images, together with the local SAR-retrieved wind direction. Therefore, several C-band models (CMOD IFR2, CMOD4, and CMODS) are available, which were developed for VV polarization, and have to be extended for HH polarization. To compare the different algorithms and C-band models as well as demonstrate their applicability, SAR-retrieved wind fields are compared to numerical-model results considering advanced SAR (ASAR) data from Environmental Satellite (ENVISAT), a European satellite.
The global availability of synthetic aperture radar (SAR) wave mode data from the European Remote Sensing (ERS) satellites ERS-1 and ERS-2, as well as ENVISAT, allows for the investigation of the wind field over the ocean on a global and continuous basis. For this purpose, 27 days of ERS-2 SAR wave mode data were processed, representing a total of 34310 imagettes of size 10 km ×5 km, available every 200 km along the satellite track. In this paper, two methods for retrieving wind speeds from SAR imagettes are presented and validated, showing the applicability of ENVISAT alike SAR wave mode data for global ocean wind retrieval. The first method is based on the well-tested empirical C-band scatterometer (SCAT) models, which describe the dependency of the normalized radar cross section (NRCS) on wind speed and direction. To apply C-band models to SAR data, the NRCS needs to be accurately calibrated. This is performed by a new efficient method utilizing a subset of colocated measurements from ERS-2 SCAT and model winds from the European Centre for Medium-Range Weather Forecast (ECMWF). SAR wind speeds are computed from the calibrated imagettes and compared to the entire set of colocated ERS-2 SCAT and ECMWF model data. Comparison to ERS-2 SCAT winds result in a correlation of 0.95 with a bias of -0.01 m s<sup>-1</sup> and an rms error of 1.0 m s<sup>-1</sup>. The second approach is based on neural networks (NNs), which allow the retrieval of wind speeds from uncalibrated SAR imagettes. NNs are trained using the mean intensity of ERS-2 SAR imagettes and colocated wind data from the ERS-2 SCAT and ECMWF model data. Validation of the NN-retrieved SAR wind speeds to ERS-2 SCAT and ECMWF model wind data result in a correlation of 0.96 with a bias of -0.04 m s<sup>-1</sup> and an rms error of 0.93 m s<sup>-1</sup>.
A mathematical model for non-Rayleigh microwave sea echo is developed which describes explicitly the dependence of statistical properties of the radar cross section on the area of sea surface illuminated by the radar. In addition to the first probability distribution of the scattered radiation, its temporal and spatial correlation functions are also considered. It is shown that, in general, these correlation functions decay on at least two scales, the second, non-Rayleigh, contributions being strongly dependent on the properties of a "single scatterer." Predictions of the model are found to be in qualitative agreement with existing experimental data. A new class of probability distributions, the " K -distributions," is introduced, which may prove useful for fitting such data.
The complex amplitude of scattered radiation at a point is modeled as a superposition of random amplitude, random phases, and random number of multipaths. When the phases have uniform probability density and the multipaths are governed by a negative binomial distribution, it is found that the probability density of the intensity of the scattered radiation becomes K distributed as the average number of multipaths tends to infinity. This density function has the property that its normalized second moment is always greater than two, though measurements have shown that the normalized second moment can lie below two but must be greater than or equal to one. The paper is devoted to a generalization in which the phases are nonuniformly distributed but the multipaths are still governed by the negative binomial distribution. In the limiting case, in which the average number of multipaths tends to infinity, the probability density of the scattered intensity is shown to be a generalization of the K-density function. This function has the property that its second moment is greater than or equal to unity. In addition, the moments of the scattered intensity are evaluated when the average number of multipaths is finite.
The generalized K distribution, introduced recently by Barakat [J. Opt. Soc. Am. A 3, 401 (1986)] as a model for the intensity of radiation scattered with a nonuniform distribution of phase by a fluctuating population of objects, is derived on the bases of simpler and more general arguments. In particular, Barakat’s analysis is based on a random walk in two dimensions and a specific model for the phase distribution; our analysis is presented in terms of an n-dimensional walk with a directional bias that, apart from its scaling property as a function of the mean number of scatters, is arbitrary. The relationship of the generalized K model with possible underlying scattering mechanisms is also discussed.
Due to its large coverage, satellite based SAR (Synthetic Aperture Radar) has proven very useful in large scale oil spill detection and monitoring. However, radar backscatter values from oil spills are not unique. Other phenomena may dampen out the capillary and short gravity waves which are sensed by the radar, and cause dark areas (generally called slicks) in the SAR imagery. At present, operational satellite SAR oil spill detection systems are based on an operator's experience in slick classification. Discrimination between oil and lookalikes includes analysis of shape, size,dB-values, gradients, texture, instantaneous wind, currents, platform-, ship lane- and natural seepage locations. However, data showing the temporal evolvement of the wind (i.e. the wind history) during the day(s) before SAR imaging, may also contribute significantly in the interpretation process. The geometrical shape of the slick is matched against wind history to estimate slick age. A procedure for such use of wind history in SAR oil spill detection is described by simple flow charts, and some examples are discussed. The method is in practical use at Tromso Satellite Station.
When the number of steps in a random walk varies, the distribution of the resultant vector components in the limit of large mean step number may be non-Gaussian. The statistics and temporal correlation properties of one class of such non-Gaussian limit distributions are derived and some of its potential applications are reviewed briefly.
This paper outlines a method for simulating the ocean surface as
seen by a space borne synthetic aperture radar (SAR). A statistical
approach was chosen in order to integrate the different mechanisms
responsible for the imaging of the surface waves. A 2D sea-state surface
was simulated, then divided in facets (of approximately the size of a
resolution cell) and each facet was given a statistical behaviour under
a surface constraint. Different parameters may be adjusted, such as
swell direction or sea-state roughness. Tests and measurements developed
for the analysis of SAR images are then used in an effort to assert the
characterization of the sea surface under various conditions of
wavelength and roughness
This paper presents the state of the art for oil spill detection in the world oceans. We discuss different satellite sensors and oil spill detectability under varying conditions. In particular, we concentrate on the use of manual and automatic approaches to discriminate between oil slicks and look-alikes based on pattern recognition. We conclude with a discussion of suggestions for further research with respect to oil spill detection systems.
In this paper, we deal with the problem of modeling the backscattering from sea surface for low-grazing-angle and high-resolution radar systems. Based on the electromagnetic two-scale model, we analyzed both the amplitude and frequency modulations induced on the small-scale Bragg resonant waves by the large-scale surface tilt and advection due to the swell presence. Evidence of sea-clutter nonstationarity has been verified and the relationship between the variations of clutter spectral features, such as texture, Doppler centroid, and bandwidth, have been studied by processing real sea-clutter data recorded by the IPIX radar of McMaster University, Hamilton, ON, Canada. An autoregressive nonstationary process is proposed and validated to model the physical phenomenon.
The authors have studied the significance of non-Gaussian signal
statistics in some synthetic aperture radar (SAR) images of the ocean
surface. The study consisted of calculating the amplitude histogram of
the returned echoes from the images and comparing these with the
Rayleigh- and Kν-distributions, corresponding to the Gaussian and
non-Gaussian statistics, respectively. The images used were some C-band
SAR data from the Canadian airborne SAR collected during the NORCSEX'88
campaign and some ERS-1 data collected during the NORCSEX'91 campaign.
The analysis of the NORCSEX'88 data included studies of the dependency
of the signal statistics on incidence angle and meteorological and
imaging conditions. It was found, specifically at small incidence
angles, that there was a significant deviation from Gaussian statistics.
It was also found that when the wind was blowing against the waves, the
deviation from Gaussian statistics was more pronounced than when the
wind was blowing in the same direction as the waves were propagating.
The study also showed a correlation between the signal statistics and
the width of the SAR image spectra. At low incidence angles, this agrees
with the interpretation that non-Gaussian statistics may be related to
strong widebanded scattering events. However, since non-Gaussian
statistics also were observed at incidence angles as high as 50°, it
is evident that the modulation of the scattering cross section by the
long waves is also an important factor. In addition, the analysis of the
ERS-1 data showed that to account for the width of the SAR image
spectra, an azimuth smearing term, due to short scene coherence time,
had to be included. This was in the present work done by modeling the
short-coherence-time-smearing as a Gaussian low-pass filter. By this
procedure, the authors were able to obtain realistic estimates for the
average scene coherence time of the SAR scenes
An automated algorithm intended for operational use is developed
and tested for estimating wind speed and direction using ERS-1 SAR
imagery. The wind direction comes from the orientation of low frequency,
linear signatures in the SAR imagery that the authors believe are
manifestations of roll vortices within the planetary boundary layer. The
wind direction thus has inherently a 180° ambiguity since only a
single SAR image is used. Wind speed is estimated by using a new
algorithm that utilizes both the estimated wind direction and σ
<sub>0</sub> values to invert radar cross section models. The authors
show that: 1) on average the direction of the roll vortices signatures
is approximately 11° to the right of the surface wind direction and
can be used to estimate the surface wind direction to within
±19° and 2) utilizing these estimated wind directions from
the SAR imagery subsequently improves wind speed estimation, generating
errors of approximately ±1.2 m/s, for ERS-1 SAR data collected
during the Norwegian Continental Shelf Experiment in 1991
In this paper, we present a new stochastic model to characterize the electromagnetic field in reverberating chambers. The model is physically based and is capable to embody as special cases the reference thermodynamic limit (perfectly stirred case) and the Rician case (presence of unstirred component). The stochastic field model is characterized by the generalized K distribution, i.e., a three parameters probability density function (pdf) that ensures a continuous and physically consistent transition among very different scattering scenarios. Theoretical modeling is supported by means of an experimental study based on measurements made at the Universita` di Napoli Parthenope, Naples, Italy, formerly Istituto Universitario Navale (IUN), mode-stirred reverberating chamber.
Reverberating chambers are of recognized relevance in many EMC
applications since they allow us to generate a reference electromagnetic
statistical field. This paper investigates the electromagnetic field in
reverberating chambers. Reverberating chambers are large overmoded
cavities wherein a nonstationary electromagnetic field is present. This
electromagnetic field is stochastic in nature; therefore, proper
analytical tools need to be considered in order to characterize it. We
deal with the theoretical aspects related to the characterization of the
electromagnetic field inside the chamber and with the application to the
case of the reverberating chamber of the Istituto Universitario Navale
Principles of Ocean Physics
- J R Apel
J. R. Apel, " Principles of Ocean Physics ". London, Academic Press: International Geophysics Series, 1990.
non-fully developed sea state characteristics from real aperture radar remote rensing
- lemaire
D. Lemaire, " Non-fully Developed Sea State Characteristics from Real Aperture Radar Remote Rensing, " PhD Thesis, Université Catholique de Louvain, Louvain-la-Neuve, Belgium, 220 pp., 1998.
marine oil pollution: technologies and technologies for detection and early warning
- delilah
H. A. Delilah, " Marine Oil Pollution: Technologies and Technologies for Detection and Early Warning, " European Commission JRC report, EUR 20231 EN, 2002.
Global Wind Speed Retrieval from SAR
- J Horstmann
- H Schiller
- J Schultz-Stellenfelth
- S Lehner