Spectral Library for Oil Types

Oil spill events are a crucial environmental issue. Although they are adequately detected using SAR images, there is a weakness regarding oil type and thickness detection. Thus, potentials of hyperspectral methods in this field are also investigated. Towards this direction, spectral signatures of different oil types for various thicknesses and in various time intervals are very useful, since they may serve as endmembers in unmixing and classification models. This paper describes the development of a relevant spectral library using spectral measurements with a GER1500 spectroradiometer. Results showed that spectral signatures of different oils types present distinctive characteristics through the whole spectral region. These characteristics are useful for oil type discrimination. Relatively high correlation has been observed between the reflectance values and the oil spill thickness, as well as the age of the oil spills. Consequently, best-fit functions, which effectively express the relationship between the reflectance and oil spill thickness as well as the age of the oil spills, have been calculated. According to these results, there is high potential of using hyperspectral imagery in order to estimate oil spill features adequately.