Oil spill detection on the ocean surface using hybrid polarimetric SAR imagery

Hybrid-polarimetric SAR (synthetic aperture radar) is a new SAR mode, with relatively simple architecture, low cost, and wide swath, which will be carried by several Earth-observing systems from now to the near future. Here, we show how the second Stokes parameter of hybrid-polarimetric SAR can be employed to detect oil on the ocean surface using the classic well-known Otsu threshold methodology, in relation to contributions from different polarizations and dampening effects on backscatter intensity, neglecting the specific scattering mechanisms and oil types for an oil-covered surface. The detection methodology is demonstrated to be reliable in three example cases: oil-on-water experiments conducted by the Norwegian Clean Seas Association, natural oil seeps from the Gulf of Mexico, and observations from the Deep Water Horizon oil spill disaster in 2010.     

结合变差纹理特征的极化SAR建筑物震害信息提取

快速评估建筑物地震灾害信息对地震应急救援工作有着指导意义,而极化SAR具有全天候、全天时的特点,因此利用极化SAR图像提取震害信息已逐渐成为研究热点。虽然极化SAR具有丰富的极化信息,但其纹理信息不可忽略,尤其是完好的人工建筑物在图像上呈现规则的纹理特征,而倒塌建筑区域纹理分布杂乱,因此结合纹理信息也可以很好地提取建筑物信息。以2010年玉树地区的全极化SAR数据为研究对象,首先,利用Yamaguchi分解的体散射分量PV提取了SAR图像中的建筑物区域以及道路、水系等非建筑物信息,在此基础上,对相干散射矩阵T11分量中倒塌建筑物、完好建筑区域进行变差计算,根据变差曲线确定变程a后,再对建筑物区域采取窗口m*m(m=3*a)进行变差计算得到变差纹理信息,最后利用FCM算法对变差纹理信息分别提取完好建筑物和倒塌建筑物区域,为了对比分析,文章利用Yamaguchi分解的二次散射分量PD提取完好建筑物区域,与震后光学遥感图像对应样本点进行人工验证,得到完好建筑物的提取精度为80.18%,倒塌建筑物的提取精度为84.54%,道路水系的提取精度为77.58%。     

SAR observation and model tracking of an oil spill event in coastal waters

Oil spills are a major contributor to marine pollution. The objective of this work is to simulate the oil spill trajectory of oil released from a pipeline leaking in the Gulf of Mexico with the GNOME (General NOAA Operational Modeling Environment) model. The model was developed by NOAA (National Oceanic and Atmospheric Administration) to investigate the effects of different pollutants and environmental conditions on trajectory results. Also, a Texture-Classifying Neural Network Algorithm (TCNNA) was used to delineate ocean oil slicks from synthetic aperture radar (SAR) observations. During the simulation, ocean currents from NCOM (Navy Coastal Ocean Model) outputs and surface wind data measured by an NDBC (National Data Buoy Center) buoy are used to drive the GNOME model. The results show good agreement between the simulated trajectory of the oil spill and synchronous observations from the European ENVISAT ASAR (Advanced Synthetic Aperture Radar) and the Japanese ALOS (Advanced Land Observing Satellite) PALSAR (Phased Array L-band Synthetic Aperture Radar) images. Based on experience with past marine oil spills, about 63.0% of the oil will float and 18.5% of the oil will evaporate and disperse. In addition, the effects from uncertainty of ocean currents and the diffusion coefficient on the trajectory results are also studied.     

Ocean wind fields retrieved from the advanced synthetic aperture radar aboard ENVISAT

In this paper an algorithm is presented which enables high-resolution ocean surface wind fields to be retrieved from the advanced synthetic aperture radar (ASAR) data acquired by the European remote sensing satellite ENVISAT. Wind directions are extracted from wind-induced streaks that are visible in ASAR images at scales above 200 m and that are approximately in line with the mean surface wind direction. Wind speeds are derived from the normalized radar cross section (NRCS) and image geometry of the calibrated ASAR images, together with the local ASAR-retrieved wind direction. Therefore the empirical C-band model CMOD4, which describes the dependency of the NRCS on wind and image geometry, is used. CMOD4 is a semi-empirical model, which was originally developed for the scatterometer of the European remote sensing satellites ERS-1 and 2 operating at C-band with vertical polarization. Consequently, CMOD4 requires modification when applied to ASAR images that were acquired with horizontal polarization in transmitting and receiving. This is performed by considering the polarization ratio of the NRCS. To demonstrate the applicability of the algorithm, wind fields were computed from several ENVISAT ASAR images of the North Sea and compared to atmospheric model results of the German weather service.Acknowledgements The authors were supported by the German Bundesministerium für Bildung und Forschung (BMBF) in the framework of the project. A new perspective of the Ocean ENVISAT Oceanography (ENVOC). The ENVISAT ASAR data were kindly made available by the European Space Agency in the framework of the ENVISAT Project AO-ID 220, Biological and geophysical parameters from synthetic aperture radar over the ocean (BIGPASO).     

Determining the oil weighting factor of oil-water emulsion based on the fractional Weierstrass scattering model

The formation of oil-water emulsion often occurs when oil is spilled into the ocean. Oil weighting factor of oil-water emulsion is one of the most important parameters for emergent oil-spill microwave monitoring. A new method is proposed here to evaluate the oil weighting factor based on fractional Weierstrass scattering model. By using the proposed method, we analyze the Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR) L-band fully polarimetric data acquired during 2010 Deepwater Horizon oil spill disaster event in the Gulf of Mexico. The result shows that our method performs well in evaluating oil weighting factor of oil-covered area.     

基于全极化Radarsat-2影像的建筑物震害评估

利用震后灾区全极化SAR影像可快速提取建筑物震害信息,为应急救援的快速有效实施提供重要的灾情信息支持。本文建立基于极化散射矩阵的Pauli-Wishart监督极化分类的建筑物震害信息提取和以街区为尺度的震害程度评估方法,包括Pauli分解、Wishart监督分类和遥感震害指数提取,并利用玉树县城区2010年4月14日青海玉树7.1级地震震后全极化Radarsat-2影像,提取了建筑物震害信息。经统计,确定结果总体分类精度达到0.81,Kappa值为0.61,表明本文提取建筑物震害的方法是可行的。     

Development of airborne oil thickness measurements

A laboratory sensor has now been developed to measure the absolute thickness of oil on water slicks. This prototype oil slick thickness measurement system is known as the laser-ultrasonic remote sensing of oil thickness (LURSOT) sensor. This laser opto-acoustic sensor is the initial step in the ultimate goal of providing an airborne sensor with the ability to remotely measure oil-on-water slick thickness. The LURSOT sensor employs three lasers to produce and measure the time-of-flight of ultrasonic waves in oil and hence provide a direct measurement of oil slick thickness. The successful application of this technology to the measurement of oil slick thickness will benefit the scientific community as a whole by providing information about the dynamics of oil slick spreading and the spill responder by providing a measurement of the effectiveness of spill countermeasures such as dispersant application and in situ burning.

This paper will provide a review of early developments and discuss the current state-of-the-art in the field of oil slick thickness measurement.     

Phase distribution of hydrocarbons in the water column after a pelagic deep ocean oil spill

Spills from wrecks are a potential major source of pollution in the deep ocean. However, not much is known about the fate of a spill at several kilometers depth, beyond the oceans continental shelves. Here, we report the phase distribution of hydrocarbons released from the wrecks of the Prestige tanker, several years after it sank in November 2002 to depths between 3500 and 3800 m. The released oil reached the surface waters above the wrecks without any signs of weathering and leaving an homogenous signature throughout the water column. At depths of several kilometers below the sea surface, the occurrence and spread of the deep sea oil spill could be evaluated better by quantifying and characterizing the dissolved hydrocarbon signature, rather than just the investigation of hydrocarbons in the suspended particulate matter.     

Modeling oil spill trajectory in coastal waters based on fractional Brownian motion

This paper proposes a numerical method to simulate oil spill trajectories, which are affected by the combination of advection, turbulent diffusion and mechanical spreading process, based on a particle tracking algorithm. Recent studies have shown that the trajectories of drifters on the ocean surface have a fractal structure that is far from being described using ordinary Brownian motion. Thus, in modeling the diffusion process, a discrete method has been employed for the generation of fractional Brownian motion (fBm) to illustrate superdiffusive transport. The algorithm is implemented to predict oil slick trajectories following the “Arteaga” oil spill accident that occurred near the Dalian coastal region in 2005. When compared with the observed data and the results of traditional diffusion modeling, the numerical results based on the fBm model are encouraging.     

Detection of macroalgae blooms by complex SAR imagery

Increased frequency and enhanced damage to the marine environment and to human society caused by green macroalgae blooms demand improved high-resolution early detection methods. Conventional satellite remote sensing methods via spectra radiometers do not work in cloud-covered areas, and therefore cannot meet these demands for operational applications. We present a methodology for green macroalgae bloom detection based on RADARSAT-2 synthetic aperture radar (SAR) images. Green macroalgae patches exhibit different polarimetric characteristics compared to the open ocean surface, in both the amplitude and phase domains of SAR-measured complex radar backscatter returns. In this study, new index factors are defined which have opposite signs in green macroalgae-covered areas, compared to the open water surface. These index factors enable unsupervised detection from SAR images, providing a high-resolution new tool for detection of green macroalgae blooms, which can potentially contribute to a better understanding of the mechanisms related to outbreaks of green macroalgae blooms in coastal areas throughout the world ocean.     

Ocean wave imaging mechanism by imaging radar

Analytical representations of the high frequency spectra of ocean wave and its variation due to the variation of ocean surface current are derived from the wave-number spectrum balance equation. The ocean surface imaging formulation of real aperture radar (RAR) is given using electromagnetic wave backscattering theory of ocean surface and the modulations of ocean surface winds, currents and their variations to RAR are described. A general representation of the phase modulation induced by the ocean surface motion is derived according to standard synthetic aperture radar (SAR) imaging theory. The detectability of ocean current and sea bottom topography by imaging radar is discussed. The results constitute the theoretical basis for detecting ocean wave fields, ocean surface winds, ocean surface current fields, sea bottom topography, internal wave and so on.     

Analysis of the reliability of a statistical oil spill response model

A statistical oil spill response model is developed and validated by means of actual oil slick observations reported during the Prestige accident and trajectories of drifter buoys. The model is based on the analysis of a database of hypothetical oil spill scenarios simulated by means of a Lagrangian transport model. To carry out the simulations, a re-analysis database consisting of 44-year hindcast dataset of wind and waves and climatologic daily mean surface currents is used. The number of scenarios required to obtain statistically reliable results is investigated, finding that 200 scenarios provide an optimal balance between the accuracy of the results and the computational effort. The reliability of the model was analyzed by comparing the actual data with the numerical results. The agreement found between actual and numerical data shows that the developed statistical oil spill model is a valuable tool to support spill response planning.     

Automatic Synthetic Aperture Radar based oil spill detection and performance estimation via a semi-automatic operational service benchmark

Today the health of ocean is in danger as it was never before mainly due to man-made pollutions. Operational activities show regular occurrence of accidental and deliberate oil spill in European waters. Since the areas covered by oil spills are usually large, satellite remote sensing particularly Synthetic Aperture Radar represents an effective option for operational oil spill detection. This paper describes the development of a fully automated approach for oil spill detection from SAR. Total of 41 feature parameters extracted from each segmented dark spot for oil spill and ‘look-alike’ classification and ranked according to their importance. The classification algorithm is based on a two-stage processing that combines classification tree analysis and fuzzy logic. An initial evaluation of this methodology on a large dataset has been carried out and degree of agreement between results from proposed algorithm and human analyst was estimated between 85% and 93% respectively for ENVISAT and RADARSAT.     

Some aspects of the Castillo de Bellver oilspill

The Castillo de Bellver incident resulted in the loss of between 160 000 and 190 000 t of crude oil. Not only was this the largest spill to date in South Africa, but it occurred near an area of great ecological sensitivity, and in the vicinity of important commercial fishing grounds. Despite this, observations during the spill and on subsequent surveys indicated that environmental damage was minimal. This can only be attributed to the favourable direction of the wind, which caused the slick to move offshore and into the Benguela current, thus allowing for its natural dispersion. Some concern remains over oil still trapped in the sunken stern section.     

Temporal variation of chlorophyll a concentration in the coastal waters affected by the Hebei Spirit oil spill in the West Sea of Korea

Time series changes in chlorophyll a concentration before and after the Hebei Spirit oil spill that occurred in December 2007 were analyzed using NCEP wind and SeaWiFS/MODIS ocean color data. Prevailing southwesterly winds and northeast/southwestward tidal currents pushed the oil towards Korea’s West Sea coast of Taean. After the oil spill, daily chlorophyll a concentration decreased about 45-50% compared to the normal condition before the oil spill, and this decrease continued for about two weeks. Monthly mean chlorophyll a concentration in December 2007 was lower compared to the average value for the same month between 1998 and 2007, but, in October and November 2007 before the spill and in January-February 2008 after the spill, the concentration value was higher than average for the same period between 1998 and 2007.     

Oil slicks on water surface: Breakup,coalescence, and droplet formation under breaking waves

The ability to calculate the oil droplet size distribution (DSD) and its dynamic behavior in the water column is important in oil spill modeling. Breaking waves disperse oil from a surface slick into the water column as droplets of varying sizes. Oil droplets undergo further breakup and coalescence in the water column due to the turbulence. Available models simulate oil DSD based on empirical/equilibrium equations. However, the oil DSD evolution due to subsequent droplet breakup and coalescence in the water column can be best represented by a dynamic population model. This paper develops a phenomenological model to calculate the oil DSD in wave breaking conditions and ocean turbulence and is based on droplet breakup and coalescence. Its results are compared with data from laboratory experiments that include different oil types, different weathering times, and different breaking wave heights. The model comparisons showed a good agreement with experimental data.     

Coseismic deformation and slip distribution of the 1997 7.5 Manyi, Tibet, earthquake from InSAR measurements

We use interferometric synthetic aperture radar (InSAR) observations to investigate the coseismic deformation and slip distribution of the 1997 M_w7.5 Manyi earthquake, a left-lateral strike-slip earthquake occurred on the west portion of the Kunlun fault in the northern Tibet, China. The fault trace is constrained by the combination of interferometric coherence image and azimuth offset image. The total length of the identified fault is about 170 km. We estimate the source parameters using a seven-segment fault model in a homogeneous elastic half-space. We first use a uniform slip model to estimate the slip, width, dip and rake for each segment, resulting in a maximum slip of 5.5 m with a depth of 11 km on the fourth segment. The average dip of the uniform slip model is about 93° northward and the average rake is about −2°. We then use a distributed slip model to estimate the pure strike-slip and oblique slip distribution, respectively. In the distributed slip model, the fault plane is discretized into 225 patches, each of them 4 km × 4 km. We fix the optimal geometric parameters and solve for the slip distribution using a bounded variable least-squares (BVLS) method. We find a geodetic moment of 1.91 × 10²⁰ Nm (M_w7.5), of which almost 68% released in the uppermost 8 km and 82% in the uppermost 12 km. For all the models used in this study, the synthetic profiles along strike show asymmetric displacements on the opposite sides of the fault, which are in agreement with the observations. This suggests that a linear elastic model with variable and non-vertical dips is also reasonable for the mechanism of the Manyi earthquake.     

Comparing the performance of a 2‐D finite element and a 2‐D finite volume model of floodplain inundation using airborne SAR imagery

The performances of a finite volume model (SFV) and finite element model (TELEMAC‐2D) in reproducing inundation on a 16 km reach of the river Severn, United Kingdom, are compared. Predicted inundation extents are compared with 4 airborne synthetic aperture radar images of a major flood event in November 2000, and these are used to calibrate 2 values of Manning's n for the channel and floodplain. The four images are shown to have different capacities to constrain roughness parameters, with the image acquired at low flow rate doing better in determining these parameters than the image acquired at approximately peak flow. This is assigned to the valley filling nature of the flood and the associated insensitivity of flood extent to changes in water level. The level of skill demonstrated by the models, when compared with inundation derived using a horizontal water free surface, also increases as flow rate drops. The two models show markedly different behaviours to the calibration process, with TELEMAC showing less sensitivity and lower optimum values for Manning's n than SFV. When the models are used in predictive mode, calibrated against one image and predicting another, SFV performs better than TELEMAC. Copyright © 2007 John Wiley & Sons, Ltd.     

Analysis of waves observed by synthetic aperture radar across ocean fronts

Ocean Dynamics - In this study, synthetic aperture radar (SAR) imaging of waves across ocean fronts was investigated using C-band Sentinel-1 VV-polarized SAR imagery collected over the Yangtze and...     

Vertical mixing of oil droplets by breaking waves

Oil spilled on a sea surface can be dispersed by a variety of natural processes, of which the influence of breaking waves is dominant. Breaking waves are able to split the slick into small droplets, facilitating oil mixing in the water column. Vertical dynamics of the droplets plays a major role in the oil mass exchange between the slick and the water column. In this paper a mathematical model of oil droplet mixing by breaking waves is developed. The model uses a kinetic approach to describe the vertical exchange of the droplets at the interface between the slick and the water column. The majority of the coefficients and parameters are conveniently combined into a single "mixing factor". The model is verified using sensitivity analysis and empirical formulae of other authors. The model permits a rapid estimation of the amount of dispersed oil under the breaking waves. The ultimate goal of the research is to parameterise influence of breaking waves on vertical mixing of oil droplets to be used in a general 3-D oil spill model.