Crude oil, petroleum product, and water discrimination on terrestrial substrates with airborne imaging spectroscopy

摘要

The Deepwater Horizon explosion and subsequent sinking produced the largest oil spill in U.S. history. One of the most prominent portions of the response is mapping the extent to which oil has reached thousands of miles of shoreline. The most common method of detecting oil remains visual spotting from airframes, supplemented by panchromatic / multispectral aerial photography and satellite imagery. While this imagery provides a synoptic view, it is often ambiguous in its ability to discriminate water from hydrocarbon materials. By employing spectral libraries for material identification and discrimination, imaging spectroscopy supplements traditional imaging techniques by providing specific criteria for more accurate petroleum detection and discrimination from water on terrestrial backgrounds. This paper applies a new hydrocarbon-substrate spectral library to SpecTIR HST-3 airborne imaging spectroscopy data from the Hurricane Katrina disaster in 2005. Using common material identification algorithms, this preliminary analysis demonstrates the applicability and limitations of hyperspectral data to petroleum/water discrimination in certain conditions. The current work is also the first application of the petroleum-substrate library to imaging spectroscopy data and shows potential for monitoring long term impacts of Deepwater Horizon.