Distinguishing Mineral Oil Slicks From Low-Wind Areas Using High-SNR Synthetic Aperture Radar

A method for distinguishing marine oil slicks from radar-dark, low wind areas in open water using single acquisitions of high-SNR SAR imagery is reported. The study uses data acquired by DLR’s airborne F-SAR X- and L-band SAR instrument during a controlled release, oil-on-water campaign, as well as NASA’s UAVSAR L-band SAR instrument imaging the Aleutian Islands in Alaska and the Coal Oil Point seep field near Santa Barbara, California. Time-series of images from three different campaigns are analyzed, all containing either verified oil slicks or low-wind zones. The product we propose, which we refer to as KS index maps, is based on the Kolmogorov-Smirnov test statistic and is used as a measure of separability between ensembles of pixels inside vs. outside a radar-dark area. We use the theoretical Polarimetric Two-Scale Model to predict the radar returns from oil slicks and low-wind areas relative to those from nearby open water as the basis of our method. In the resulting maps, KS index values for oil slick are consistently between -1 and 0 and KS index values for low-wind zones are consistently between 0 and +1. Noise analysis of all imagery revealed that high-SNR data are essential for accurately classifying radar-dark zones; otherwise, negative KS index values are obtained regardless of whether the area is a low-wind zone or an oil slick. The method is sufficiently simple and generic to be applied in near-real-time and without special processing code and can be easily implemented to aid in cleanup operations.