Author
|
Submitted to: American Society of Civil Engineers
Publication Type: Peer Reviewed Journal Publication Acceptance Date: 3/6/2000 Publication Date: 6/20/2000 Citation: Clark, P.E., Stream shading assessment with airborne multispectral videography., p.1-11., 2000 conference Interpretive Summary: Heat is a major pollutant of stream waters on Western rangelands. To comply with the Federal Clean Water Act of 1972, Total Maximum Daily Load (TMDL) allocations for heat or elevated stream temperature are being developed for stream systems throughout the Western states. Monitoring compliance to water quality standards is currently done using ground-based methods which are inadequate for the task. A remote sensing technique is needed to practically and efficiently provide stream temperature information for entire stream systems. Stream shading by riparian vegetation can strongly influence stream temperature by reducing solar heating. Estimation of stream shading using remotely-sensed parameters such as the canopy structure of riparian vegetation has been suggested as a possible surrogate for direct stream temperature measurement. In this study, supervised classification of airborne multispectral videography was evaluated as means of assessing the canopy structure of riparian vegetation within Reynolds Creek Experimental Watershed in southwestern Idaho. Multispectral videography of riparian vegetation was classified into 2 classes; mature trees and shrubs/juvenile trees, with 85% overall user accuracy using a maximum likelihood classifier. Additionally, application of the Soil Adjusted Vegetation Index (SAVI) to multispectral videography was qualitatively evaluated as a means of assessing the greenness or canopy complexity of riparian vegetation. The SAVI values tended to closely follow trends in the canopy complexity of riparian vegetation. These preliminary results suggest analysis of airborne multispectral videography may provide key inputs useful for modeling stream shading and estimating stream temperature in rangeland systems. Technical Abstract: Heat is a major pollutant of stream waters on Western rangelands. To comply with the Federal Clean Water Act of 1972, Total Maximum Daily Load (TMDL) allocations for heat or elevated stream temperature are being developed for stream systems throughout the Western states. Monitoring compliance to water quality standards is currently done using ground-based methods which are inadequate for the task. A remote sensing technique is needed to practically and efficiently provide stream temperature information for entire stream systems. Stream shading by riparian vegetation can strongly influence stream temperature by reducing solar heating. Estimation of stream shading using remotely-sensed parameters such as the canopy structure of riparian vegetation has been suggested as a possible surrogate for direct stream temperature measurement. In this study, supervised classification of airborne multispectral videography was evaluated as means of assessing the canopy structure of riparian vegetation within Reynolds Creek Experimental Watershed in southwestern Idaho. Multispectral videography of riparian vegetation was classified into 2 classes; mature trees and shrubs/juvenile trees, with 85% overall user accuracy using a maximum likelihood classifier. Additionally, application of the Soil Adjusted Vegetation Index (SAVI) to multispectral videography was qualitatively evaluated as a means of assessing the greenness or canopy complexity of riparian vegetation. The SAVI values tended to closely follow trends in the canopy complexity of riparian vegetation. These preliminary results suggest analysis of airborne multispectral videography may provide key inputs useful for modeling stream shading and estimating stream temperature in rangeland systems. |
