On direct extraction of scene component fractions and crown cover distribution in open forest canopies using high spatial resolution winter imagery

Spatial patterns of understory and overstory constitute an as-yet poorly un derstood obstacle to the derivation of boreal forest structural variables t hrough remotely-sensed image data. The high resolution reflectance imagery collected seasonally as part of the Boreal Eco-system and Atmosphere Study (BOREAS) presents an opportunity to increase overall understanding of this complicated interplay. A research project is currently underway to compare and contrast two methods of separating understory from overstory in this im agery, one "directly" through classification and the other "indirectly" thr ough end-member analysis of a high spectral resolution data set. Winter ima gery is investigated to assess the best method(s) of classification into cr own and snow understory. Simple thresholds are considered, followed by unsu pervised and supervised classification tests by variation of scene type, tr ee-type, number of classes and number of wavelength channels. K-Means (Iter ative Optimization) Clustering reveals that less than four classes is insuf ficient to effectively separate the land-cover classes for both tree-types investigated and that more than four classes leads to trivial division of e xisting classes. Results from this unsupervised classification are used to build training areas for Maximum-Likelihood supervised classifications. JM (Bhattacharya) distances are calculated, and it is shown that three channel s provide optimum separability for both tree-types. A working methodology f or winter image classification is proposed.