Influx of hydrocarbon contaminants to the coasts of United Arab Emirates is a serious environmental problem that threaten marine habitat. This paper presented two datasets to extract information on hydrocarbon pollution in marine environment utilizing visible to near infrared portion of the electromagnetic spectrum. The datasets used were Land-sat ETM+ images for Dubai and Fujirah coasts, and experimental visible to near infrared spectra of petroleum hydrocarbon in water. The experimental spectra were measured using GER 3700 spectroradiometer. The study applied image analyses techniques and radiance ratios on corrected Landsat ETM+ images in order to detect oil spills from tankers. The reported results demonstrated proper bands or channels with techniques useful for oil spill detection using images of Landsat ETM+. The study found that radiances extracted from visible and near infrared images are useful to assess the thickness of hydrocarbon spills. The study also found that areas of thicker oil are slightly brighter because they suppress the signal from the underlying seawater. The spectra of GER 3700 provide potentials for identification of possible contaminants as well as classification within a particular hydrocarbon category (e.g. aromatics). The presented techniques hold promise for rapid analysis of petroleum hydrocarbons in marine environment.