DIFFICULTIES IN DETERMINING MEANINGFUL SPECTRAL 
SIGNATURES OF FOREST TREE CANOPIES 
J. Vlcek 
Department of Civil Engineering 
and Faculty of Forestry, 
University of Toronto, Toronto, Ont. Canada 
ABSTRACT 
Determination of spectral reflectance curves (signatures) of forest 
tree canopies presents many difficult problems. Some of these are: need for 
elevated observation platforms, lack of suitable field spectroradiometric 
equipment, need to monitor incident irradiance, bidirectional character of re 
flectance phenomena, 3-dimensionality of forest crown canopy, biomass, topo 
graphical conditions of the terrain, site characteristics of the stand and 
others. 
This article describes an attempt to determine, in situ, spectral 
signatures of several tree species found in Gatineau Park and Larose 
Experimental Forest in the vicinity of Ottawa, Canada. The data was collected 
in August 1973. Special emphasis was placed on their replicability so that 
their statistical properties could be assessed. The results show a great 
difficulty in finding "clean" signatures that could be termed characteristic 
of the species measured. This is attributed to large statistical variation 
of the signatures and also to the fact that, toward the end of the growing 
season, the spectral differences among trees become less pronounced. 
OBJECTIVE OF STUDY 
The study was initiated and sponsored by the Forest Management 
Institute, Department of the Environment, Government of Canada. The objective 
of the study was determination of daylight reflectance spectra within approxi 
mate band of 0.4 to l.lym of several tree species grown in the forest in two 
localities near Ottawa. The trees measured were: white pine (Pinus strobus L.), 
red pine (Pinus resinosa Ait.), white spruce (Pices glauca Voss), sugar maple 
(Acer saccharum Marsh.) and trembling aspen (Populus tremuloides Michx.). 
The work began at the end of July and finished in the first week of September. 
Spectral reflectance in this study was defined by the ratio (%) 
spectral radiation reflected in the direction of the instrument 
incident spectral radiation of daylight (sun & sky) 
Spectral reflectance so defined differs from laboratory measured quantity in 
that it depends strongly on the direction of sun's irradiance and the 
direction in which the reflectance is observed. The bidirectional character 
of the reflectance can be specified by three angles: zenith angle of the sun