| The normalized difference vegetation index (NDVI) has become widely accepted because of its compensation for changing illumination conditions and simplicity (Lillesand and Kiefer, 1994). Although NDVI compensates, partially, for the effects of topography on remote sensing measurements, the question remains whether signal noise, attributed to the remaining topographic effect, may explain variability in NDVI. Greater understanding of the remaining topographic effect may yield more accurate interpretation of NDVI in sensitive ecotones. Through the examination of NDVI, solar incidence angle, and known vegetation cover for a site in Rocky Mountain National Park, this thesis explores the question of whether the calculated NDVI values are related to topography and how strong is that relationship, if any, relative to the vegetation influence. Statistical tests indicate a significant relationship between incidence angle and NDVI. In large, homogeneous areas, the "noise" of incidence angle may account for up to 6% of the NDVI signal. |
