Polarimetric radar signatures of oil slicks for measuring slick thickness

Abstract

In the event of an oil spill at sea, the most critical information needed is an estimate the amount of the spill and how it is dispersing. The thickness of the oil slick provides information needed to assess both these needs. Polarimetric synthetic aperture radar has been used to estimate the thickness of oil slicks. However, since a model that describes the measurements of a fully-polarimetric radar was not available, it was difficult to optimize this measurement technique. The two-scale, resonant scattering model is utilized to derive the backscatter cross-section of ocean surfaces that are covered by thin oil slicks. The model defines the complex scattering coefficients of the co-pol and cross-pol backscattered signals. The scattering coefficients are found as a function of the thickness of the oil slicks and its dielectric constant. The scattering matrices are converted to the corresponding Stokes matrices. The Stokes matrix is used to synthesize the backscattered signal for a variety of different polarizations. The scattering cross-section is found as a function of the ellipticity angle and the ellipse orientation angle. This information is used to train an artificial neural network to find the optimum polarization as a function of the dielectric constant and slick thickness. This artificial neural network is used to extract the slick thickness. Estimates of the resolution of the thickness measurement are also given. The effect of phase noise and speckle on the accuracy of the thickness measurement are also estimated. It is concluded that a fully polarimetric, millimeter wave radar, such as a 33 GHz SAR, is adequate for estimating the thickness of most moderately thick oil spills. >