Quantifying Percent Cover of Submerged Oil Using Underwater Video Imagery

Abstract

On November 11, 2005, the double-hulled tank barge (TIB) DBL 152 struck the submerged remains of a drilling platform that collapsed during Hurricane Rita in the northwestern Gulf of Mexico. The allision and eventual capsize of the barge caused the release of approximately 1.9 million gallons of low-API gravity slurry oil, which sank to the seafloor. Throughout the response, information on the concentration of submerged oil (volume per unit area) was sought to prioritize cleanup of multiple sunken oil patches. Numerous methods were employed to detect and assess submerged oil including, among others, divers to measure oil thickness and chain-weighted sorbent drags to delineate patch size. However, none of the techniques used were able to quantify the percent cover of oil, which emerged as a critical parameter for estimating oil concentrations. This paper presents the methodology for quantifying percent cover of submerged oil used during the long-term monitoring phase of this incident. Orthographic (downward-looking) images of the seafloor were obtained using an underwater drop camera. Full-motion video was digitally recorded using a laptop computer and video capture hardware. Imagery was georeferenced by overlaying GPS coordinates onto the streaming video using a video annotator. Still images were extracted from the video using frame-grabbing software during post-processing. Percent cover of oil was then manually determined for each image using two different methods: ocular estimation and point counts. Seafloor imagery was acquired at discrete sample points along drift transects, and when visibility allowed, continuously between sample points. Percent cover data obtained using these techniques, combined with oil thickness measurements and patch size information obtained using other methods, was used to estimate the volume of oil present within a discrete patch of submerged oil discovered in September 2006. Successful application of this technique demonstrated its viability as a cost-effective means of obtaining quantitative information about percent cover of submerged oil. Moreover, this relatively low-tech system was comprised primarily of off-the-shelf components that can be readily sourced, assembled and mobilized to a submerged oil spill response area.