OIL SPILL DETECTION BY IMAGING RADARS: CHALLENGES AND PITFALLS

摘要

Criteria for discriminating between radar signatures ofrnoil films and oil-spill look-alikes visible on syntheticrnaperture radar (SAR) images of the sea surface arerncritically reviewed. The main challenge in oil spillrndetection using SAR is to discriminate between mineralrnoil films and biogenic slicks originating from secretionsrn(exudations) of biota in the water column. The claimrnthat oil spill detection algorithms based on measuring 1)rnthe reduction of the normalized radar cross sectionrn(NRCS), 2) the differences in the geometry and shapernof the surface films, and 3) the differences in texturernhave a high success rate is questioned. Furthermore, it isrnquestioned that polarimetric SAR data are of great helprnfor discriminating between mineral oil films andrnbiogenic slicks. However, differences in the statisticalrnbehavior of the radar backscattering is expected due tornthe fact that, other than monomolecular biogenicrnsurface films, mineral oil films can form multi-layers.