 |
 |
 |
 |
|
| back | ||||
Abstract
A Hydrologic Classification of the Embudo Watershed, Northern New Mexico
Track: Water Resources
Author(s): Ryan Smith
This study focuses on the application of using hydrologic parameters in a standard unsupervised classification. This focus is important because many land classifications fall short when relaying the hydrologic attributes of land cover. Standard land classifications based on spectral response alone look only at land cover as vegetation, not as what areas are most susceptible in producing runoff. This prompts the need for an enhanced hydrologic land classification that incorporates specific hydrologic parameters of elevation, slope, aspect, and texture (surface roughness). The land cover classification of the Embudo watershed was derived and classified by a K-means unsupervised algorithm. The classification was then improved hydrologically by adding channels of elevation, slope, aspect, and texture (surface roughness). These channels, along with the spatial channels of the original classification, were classified by a K-means unsupervised classification and evaluated as to their hydrologic parameters. An evaluation of the classes was performed using a GPS survey and field evaluation of the study area. The hydrologic classes derived in this study should give a better estimate of surface runoff when compared to standard land cover classes derived from spectral bands alone. The final determination of the hydrologic land cover classification compared to that produced by a standard spectrally based land cover classification will come once incorporated into a hydrologic surface runoff model.
Ryan Smith
Landmark Geographic Services Inc.
GIS
679 East 2nd Str
Durango , CO 81301
USA
Phone: 970 375 6277
E-mail: ryans@landmarkgeo.com