Abstract
Terrain influences many ecological processes. Therefore, ecologists must consider terrain effects in sampling designs. Incomplete information on the ecological role of terrain, however, makes it difficult to account for terrain variability in sample designs. As a partial solution, we introduce a protocol that automatically stratifies complex terrain. This stratification method uses image-processing techniques to cluster digital images of slope, elevation, and aspect into homogeneous groupings. The technique is demonstrated for a stratification of complex terrain disturbed by a large forest fire. By identifying terrain complexity at several spatial resolutions, we were able to place transects that sample terrain variability and fire severity representatively across a controlled range of scales.
Original language | English (US) |
---|---|
Pages (from-to) | 433-443 |
Number of pages | 11 |
Journal | Physical Geography |
Volume | 19 |
Issue number | 5 |
DOIs | |
State | Published - Sep 1 1998 |
Keywords
- Image processing
- Sample design
- Stratification
- Terrain
ASJC Scopus subject areas
- General Environmental Science
- Atmospheric Science
- Earth and Planetary Sciences (miscellaneous)
- General Earth and Planetary Sciences