Operational impacts of Glen Canyon Dam on downstream resources of the Colorado River ecosystem have been the subject of intense study and discussions among adaptive management stakeholders that include Federal, State and tribal entities, as well as recreational businesses, environmentalists, power companies, and others ( U.S. Bureau of Reclamation 1995, Schmidt et al. 1998). The Grand Canyon Monitoring and Research Center (GCMRC within the U.S. Department of Interior) has identified the following key objectives for long-term monitoring of vital components of the ecosystem between Lakes Powell and Mead, while providing stakeholders with information for decision making: 1) lowest possible impact on critical resources; 2) highest possible return of integrated scientific information; and 3) long-term, repeatable data collection. Airborne hyperspectral image data are being considered as a possible long-term monitoring tool, which hold the promise of being able to fulfill the above requirements. The present study is a pilot project which is intended to explore the extent and variety of information, relevant for long-term monitoring, that can be extracted from images of both high spectral and spatial resolution.
The Colorado River ecosystem in Glen Canyon and Grand Canyon poses special challenges for remote monitoring with respect to logistics required to support field-based studies and relative to spatial scales at which monitoring is needed. The narrow corridor of the river is confined by tall vertical cliffs, yet provides a habitat for a great variety of both native and non-native plant species. Riparian vegetation consists of stands that often cover only very small areas. Often, these stands consist of intermixed native and non-native species, and spectral signatures of the various plants sometimes display only subtle differences. Adequate change detection mapping requires both high spectral and high spatial resolution, as well as advanced processing algorithms that can detect subtle spectral differences.
Change detection for evolving sand bars, rapids and debris fans is a critical part of the GCMRC's long-term monitoring program. Additionally, changes in the aquatic ecosystem's food base, such as abundance and distribution of algae and other benthic organisms are of great interest. Presently, the need to lower river stage for 2- 3 days during annual monitoring overflights for conventional aerial photography and traditional surface measurements introduces artificial and undesirable impacts on the ecosystem (e.g., Blinn et al. 1999). With hyperspectral imagery, we expect to map riparian vegetation, terrestrial sediment deposits, as well as sand bars and benthic organisms through some depth of water, obtaining greater compositional information than can be obtained through the use of aerial photography.