Oil pollution detection using spectral fluorescent signatures (SFS)

Abstract Marine oil spills can be highly dangerous given that the wind, waves and currents can scatter an oil spill over a wide area within just a few hours in the open sea. Remote sensing can effectively be used for detecting and monitoring of oil spills. Optical sensors provide important information for oil spill contingency planning. Remote sensing techniques have contributed a great extent to the development of oil pollution monitoring systems. However, the available detection methods, mainly designed for spaceborne and airborne long-distance inspection, are too expensive and complex to be used in an operational way by relatively unskilled personnel. In the framework of DEOSOM project (European AMPERA project), an innovative water inspection method was investigated, based on the analysis of suspicious water samples in laboratory using the spectral fluorescent signature (SFS) technique. The SFS technique is based on the fluorescence properties of the oil compounds. Different oil products were analyzed in the laboratory, using eight LED light sources with wavelengths ranging from 270 to 850 nm to excite the samples and a USB4000-FL fluorescence spectrometer to register the fluorescence spectra. Finally, an algorithm to identify oil products in polluted water samples with known substances was developed. The mean of the 16 replicas was calculated for each point and used to calculate the Euclidean distance with respect to a known sample constituted by distilled water and single oil. Finally, the percentage of similarity referred to every oil was determined.