Geospatial techniques for stream research in the southern Blue Ridge Mountains
Gardiner, Edward P
MetadataShow full item record
This research demonstrates how geographic information systems (GIS) complement stream research based on sampling habitat characteristics, water quality, and biota. The study area encompasses the Upper Little Tennessee, Tuckaseegee, Pigeon, and French Broad River basins in the Blue Ridge physiographic province in western North Carolina, U.S.A. Chapter 2 presents a methodology for constructing a spatial database of linked watershed and stream attributes and assesses the planimetric accuracy of automatically extracted watershed boundaries. Watersheds extracted from 30-m USGS Level 1 digital elevation models (DEM) larger than 250 ha have negligible errors. The mean distance between automatically derived and hand digitized watershed boundaries was comparable to the radius of a circle of the same size as a DEM pixel. In Chapter 3, a watershed sediment yield model is constructed using the Revised Universal Soil Loss Equation (RUSLE). The resolution of source data largely determines the precision of model estimates that use those data. When 90-m input data are used, 80% of the variance in model output using 30-m data are explained, but precision drops rapidly as data resolution is coarsened beyond 90 m. Modeled sediment yields (tonne yr -1 ) are statistically related to baseflow sediment yield estimates calculated from stream discharge (m 3 s -1 ) and total suspended solids (mg l -1 ) measurements. Chapter 4 examines land use and physiographic factors associated with fish assemblage shifts in the Upper Little Tennessee River basin. Large and small watersheds greater and less than 39 km 2 in area have distinct fauna. Smoky sculpin proportions (Cottus bairdi ssp), widespread taxa, and restricted range taxa are statistically related to modeled sediment yield rate (tonne km -2 yr -1 ). Chapter 5 synthesizes biological, chemical, and land use data to project the effects of future land use changes onto forecasts of stream conditions in the Upper Little Tennessee River and Cane Creek basins. This research provides a powerful suite of analytical tools and perspectives relating GIS to stream research.