Remote sensing approach to map riparian vegetation of the Colorado River Ecosystem, Grand Canyon area, Arizona

Abstract:

Riparian zones in the southwestern U.S. are usually a mosaic of vegetation types at varying states of succession in response to past floods or droughts. Human impacts also affect riparian vegetation patterns. Human- induced changes include introduction of exotic species, diversion of water for human use, channelization of the river to protect property, and other land use changes that can lead to deterioration of the riparian ecosystem. This study explored the use of remote sensing to map an iconic stretch of the Colorado River in the Grand Canyon National Park, Arizona. The pre-dam riparian zone in the Grand Canyon was affected by annual floods from spring run-off from the watersheds of Green River, the Colorado River and the San Juan River. A pixel-based vegetation map of the riparian zone in the Grand Canyon, Arizona, was produced from high-resolution aerial imagery. The map was calibrated and validated with ground survey data. A seven-step image processing and classification procedure was developed based on a suite of vegetation indices and classification subroutines available in ENVI Image Processing and Analysis software. The result was a quantitative species level vegetation map that could be more accurate than the qualitative, polygon-based maps presently used on the Lower Colorado River. The dominant woody species in the Grand Canyon are now saltcedar, arrowweed and mesquite, reflecting stress-tolerant forms adapted to alternated flow regimes associated with the river regulation.