Exploring the capabilities of TM and AVIRIS sensor images for oil spill detection in marine and ocean waters

Abstract

Oil spills are major water polluting sources. Due to its devastating effects on the sea and ocean ecosystem, detecting oil pollution in the shortest time and with the highest confidence level is necessary. Remote sensing being a suitable option, the capability of Landsat multispectral data and airborne hyper-spectral data from the AVIRIS sensor was investigated for study of the 2001 oil spill in the Gulf of Mexico. In this study, a part of the 2001 oil spill data was processed in terms of cloud spots,bad pixel and atmospheric correction. The pixel purity index was used to extract the end -members of water and oil spill and the linear spectral unmixing method was used for mapping of water from oil spills. The results show that the AVIRIS image is able to detect the type and thicknesses of oil spill, due to its ability to cover the diagnostic spectral signature of oil.
Keywords: Monitoring, oil spill, remote sensing, Landsat, AVIRIS.