An evaluation of the relative value of spectral and phenological information for tree crown classification of digital images in the eastern deciduousforest

Digitized, multitemporal, small format 35 mm aerial photography was used to determine the relative value of spectral and phenological information for tree crown classification of digital images of the Eastern Deciduous Forest . This sampling scheme was designed to simulate the high spatial resolution and multitemporal data anticipated from the next generation of land observ ing satellites. The one-hectare study site, located in the West Virginia Un iversity Experimental Forest approximately 15 km east of Morgantown, West V irginia, was photographed from a light aircraft 10 times from May to Octobe r 1997 using both true color and false color film. All negatives of the plo t were digitized, rectified, enhanced, and classified using ERDAS Imagine. Differences in the timing of phenologic events between tree species, specif ically leaf development and flushing, leaf maturity, and leaf senescence, m ade it possible to separate spectrally four deciduous tree species, namely Liriodendron tulipifera, Acer rubrum, Quercus rubra, and Quercus alba, from the surrounding vegetation in the test plot. Optimally timed photography w as required during spring and autumn to coincide with seasonal variations i n leaf phenology between deciduous species. Within the visible spectrum (40 0-700 nm), spectral differences in tree crowns were observed based at least partly on leaf ontogeny and primary leaf pigments. In contrast, near-infra red (700-900 nm) reflectance was influenced primarily by plant tissue chara cteristics (Boyer et al., 1988). However, color IR produced a wide range of foliage colors which, when combined with selected bands (multi-date) of th e visible spectrum, made a significant contribution to the separation of so me individual tree species. Photography acquired in spring and fall provide d the best data for species separation. Variable canopy illumination made d igital classification of individual trees complex.