HYPERSPECTRAL IMAGE ANALYSIS FOR OIL SPILL MITIGATION

Abstract

1. ABSTRACT During the last several decades, the Chesapeake Bay has suffered from several large spill events threatening coastal habitats and species. Chesapeake Bay’s resources remain vulnerable as they share the coastal areas wit h major interstate commerce routes, underground pipelines, extensive development, large industrial facilities, and heavy shipping traffic to the ports of Norfolk and Baltimore. In order to effectively protect communities and species in jeopardy, fast and accurate determination of oil spill hazard areas is needed, particularly if monitoring large quantities of oil spilled. Where oil-handling infrastructure is aging, this need is amplified. This research addresses remote sensing, especially Hyperspectral image analysis applicable to the Chesapeake watershed. This case study is a prototype of oil spill leaks in Patuxent River in Maryland, and the associated image analysis for detecting oil spills using hyperspectral imagery and the effect on soil, water, wetland, and vegetation contaminated by oil spill. Space-borne and airborne images, as an effective survey tool, are the main source for getting real-time data. In the event of an oil spill, this information can be retrieved in short time to help authorities to plan the quickest route to the spill and formulate an effective environmental protection plan, that could be a way to reduce damages. Hyperspectral sensor affords the potential for detailed identification of materials and better estimates of their abundance. This can eliminate the false alarms of features that have the same color and appearance of oil, such as large algae blooms or jellyfish. Such phenomena may be identified by visual interpretation as a suspected oil spill using some conventional sensors. Some other types of light fuel, such as gasoline and diesel, cannot be identified visually because of their changing appearance with time. Hyperspectral sensing can record over 200 selected wavelengths of reflected and emitted energy. With this spectral information one can exploit the spectral signature of oil and also to distinguish between different types of oil (crude or light oil). HSI observations with high spectral and spatial resolution can be used to detect oil based on the spectral signature matching to identify the level of oil contamination of polluted areas in the shoreline, which are necessary for determining proper cleaning processes. 2. MOTIVATIONS In this research some of these problems can be minimized when using more advanced methodology to identify oil spill based on the spectral signature matching not by visual interpretation of the image. New techniques such as HIS should be used in order to make the proper distinctions between oil types (crude or light oil) and to properly identify natural phenomenon; The results of this research show that with HSI spectral information, the signature of oil can be used to detect minute concentrations of hydrocarbon (crude oil) on the sea and it can distinguish between different types of oil (c rude or light oil). Part of this study also includes the monitoring of oil slicks movements, dispersion in water, and identifying spills on the shoreline. The work under progressing will emphasize the oil contaminates in wetland, soil, vegetation, and grass in Patuxent River basin in Chesapeake Bay due to oil pollution. The Linear un-mixing technique will be used to distinguish between mixing signatures of different materials polluted with oil.