LABORATORY HYPERSPECTRAL INVESTIGATIONS OF GEOMATERIALS FROM THE GREAT MIAMI RIVER AT HERITAGE PARK, COLERAIN TOWNSHIP, OHIO: PROGRESS IN DEVELOPING AN ENVIRONMENTAL LIBRARY FOR COMPLEX GEOLOGIC SETTINGS.

Hyperspectral remote sensing technology has the ability to identify objects using the reflective spectra captured in each pixel of an image. Images can be obtained from numerous airborne or spaceborne platforms. Increasingly hyperspectral images are being used in more complex environments, at higher resolutions and for broader sets of tasks. Integral to the full exploitation of this technology is development of libraries of spectra, backed by detailed materials analysis to serve the advanced geospatial analyst. A main challenge in developing these libraries is detailed materials investigations of complex systems of interest. Fluvial environments are an example of such complex systems and are of interest to hydrogeologists, engineers, ecologists as well as insurance companies, law enforcement and other entities. Obtaining detailed information regarding pollutants, objects and geological materials in fluvial systems through hyperspectral remote sensing techniques is challenging. This investigation provides some constraints on the hyperspectral properties of geomaterials as well as associated and potential objects of relevance from a major regional system, the Great Miami River at Heritage Park, Colerain Township, Ohio. Several fluvial deposit samples and clasts where collected for hyperspectral analysis and follow-on SEM and XRD composition analysis. Common fluvial sediment and clasts high in phyllosilicate and carbonate content exhibit easily distinguishable and identifying absorption features, reflectance features and overall distinctive spectral signatures, regardless of sediment size. These spectra demonstrate sharp contrast to objects or items of concern such as gasoline, diesel fuel, plastics and numerous other environmentally relevant items. A comparative analysis of the geology from other glaciated regions suggests spectra may be translatable for use in other regions of Ohio, Indiana, Illinois, New York, Wisconsin and other areas. This project demonstrates progress towards detailed hyperspectral libraries that may be of utility for numerous entities.