Extraction of Hydrocarbon Content Information by Using Hyperspectral Image at Liaodong Bay, China

Jianwei Xiao,Qingjiu Tian,Yingcheng Lu,Ling Wang,Xiaoping Qi,Baihong Wen
DOI: https://doi.org/10.1109/geoinformatics.2010.5568213
2010-01-01
Abstract:the method of hyperspectral remote sensing has great potential in the area of offshore oil and gas exploration and marine pollution investigation. Combined with Hyperion hyperspectral satellite images and the surface geochemical data of dissolved hydrocarbons at Liaodong Bay, as well as the concentration data of chlorophyll and suspended material, The distribution of hydrocarbon over Liaodong Bay are presented. This research finds that the band at 579nm, 1184nm and other Hyperion image band reflectance have a good correlation with the content of hydrocarbon. Considered the chlorophyll and suspended material spectrum information influence, the study finds that band at 579nm is the ideal hydrocarbon content estimate band and carry out the distribution chart on the Hyperion image. The model was tested to have good accuracy(the relative error is 0.11). The result shows that the hydrocarbon is mainly distributed in the middle of Liaodong Bay area, and corresponded with Bohai Bay fault zone well. In addition, the concentration of dissolved hydrocarbons and suspended material in Liaodong Bay area has high inverse correlation, indicated that the concentration of suspended material can be used indirectly to verify the presence of offshore hydrocarbon.
What problem does this paper attempt to address?