On the analysis of a time series of X-band TerraSAR-X SAR imagery over oil seepages

Abstract

A large time series of 42 dual-polarimetric co-polarized TerraSAR-X (TSX) StripMap Synthetic Aperture Radar (SAR) measurements are exploited to monitor a well-known oil seep area, i.e., the Taylor Energy site in the Gulf of Mexico. A comprehensive scattering analysis is undertaken to assess the impact of SAR imaging parameters (polarization, angle of incidence - AOI, noise floor) and environmental conditions (wind speed - WS, oil properties) on single-polarization SAR-based sea oil seep observation. The main goal of this study is to evaluate the reliability of the scattering-based information derived from the time series of TSX SAR imagery. A two-scale backscattering model is considered to give a physical framework that supports a better understanding of the effects of the above-mentioned factors. Experimental results showed that the high TSX noise floor significantly limits a reliable interpretation of the slick-free sea surface and oil seep backscattering at AOIs larger than 34 and 26, respectively, since they are contaminated by noise. Hence, it is shown that, at larger AOIs, the joint contribution of noisy SAR measurements and low oil backscattering does not result in a larger oil/sea separability. The latter is not remarkably influenced by polarization and WS, under low-to-moderate conditions. Experiments also demonstrate that, when reliable SAR measurements are available, sea oil seep backscattering is affected by the oil's damping properties more than its concentration in the water column. The time variability of the polluted area is also estimated using the time series of TSX imagery and the obtained results agree with independent analysis undertaken on the same test site.