| Remote sensing technology is an important tool for monitoring, detecting, and analyzing oil spills. Researchers have explored the use of digital imagery acquired from airborne and spaceborne platforms for monitoring oil spills and for analyzing changes to oil spill thickness and contaminated areas. However, traditional digital imagery from multispectral scanners is subject to limitations of spatial and spectral resolution.; A new type of remote sensing, known as “hyperspectral sensors,” promises to revolutionize the use of remotely sensed data for a variety of applications including mapping and monitoring oil spills by eliminating the limitations of multispectral scanners. With hyperspectral sensors, it is possible to map oil spills of different types and thicknesses, as well as to detect subtle changes in oil-contaminated wetlands such as complex contaminated wet soil and vegetation.; However, despite the great promise they offer, these sensors introduce a host of problems which must be addressed before they can be routinely used in oil spill applications. For example, statistical analysis techniques commonly used to process multispectral data are not suited to the amount and dimensionality of data present in a hyperspectral image. The large volume of data, along with the CPU-intensive algorithms required to derive information from hyperspectral data, make it difficult to extract useful information.; This dissertation describes the spectral and spatial characteristics of hyperspectral data and the potential work of these data for oil spill detection and environmental applications. The advantages and disadvantages of these data for oil spills in fresh and sea water and contaminated wetlands are discussed. Furthermore, application of the Airborne Imaging Spectrometer for Applications (AISA) and the Airborne Visible/Infrared Imagery Spectrometer (AVIRIS) data for oil spill image classification is used.; Two case studies are considered, the first focusing on the extraction of oil spill spectra and the region of interest (ROI) of oil-contaminated areas in Chesapeake Bay River, Maryland; the second is focusing on target identification for oil slick type and signature feature analysis for oil spill thickness in the sea at Santa Barbara, California. Results showed that the spectral features for oil spill on fresh water and sea water could be clearly identified; hence, that spills can be mapped. In addition, oil spill thickness and slick types at different stages could be identified by analyzing the spectral features of AVIRIS data. |
