Multi-frequency and polarimetric quantitative analysis of the Gulf of Mexico oil spill event comparing different SAR systems

Abstract Synthetic aperture radar (SAR) images were acquired over the Deepwater Horizon spill's main slick with a number of different instruments during response to the catastrophic accident in 2010. These included instruments operating in different microwave frequency bands (L-C-X band), and, for TerraSAR-X and COSMO-SkyMed, more than one acquisition system operating in the X-band. In this work, for the first time the diverse SAR data acquired over the Deepwater Horizon spill have been compared and quantitatively analyzed with the goal of determining the different capabilities of the last generation SAR systems in the analysis of this important test case. Using the nearest acquisitions in time over the same areas, backscattering coefficients and polarimetric features, when available, have been evaluated. The derived information and comparison of sensors is discussed, in particular taking into consideration the Noise Equivalent Sigma Zero (NESZ) value characterizing each instrument. Although, as expected, an increase of the damping ratio values is noted at higher frequencies, the best slick discrimination capability has been obtained with the L-band UAVSAR sensor, characterized by a more effective NESZ value, while the satellite SAR sensors are shown to have in some cases significant noise contamination to polarimetric parameters. Anomalous behavior at X-band measurements has been also registered and discussed. Through the analysis of all the available data an evaluation of the impact of low noise SAR for oil characterization in the considered context is provided.