734 
1. Variations in sun-target-sensor geometry, causing variations in surface reflectance anisotropy. 
2. Variations in spectral irradiance at the target. 
3. Variations in spectral reflectance factor at any sun-target-sensor geometry with size of field of view. 
4. Changes in canopy geometry with wind 
5. Changes in canopy reflectance caused by dew. 
3.3.1. Canopy reflectivity and stress. The factors that change when the canopy become stressed and hence affect 
the canopy's reflectance are:- 
1. The percentage of vegetation cover, 
2. The canopy geometry (size, shape and orientation of the leaves) 
Again these are inter-related (Pinter et al., 1985 (33), Holben et al., 1983 (53). 
Healthy and stressed plants can often be distinguished from differences in leaf area and foliage density . These 
differences occur because stressed plants may either lose foliage or have stunted growth. As foliage density 
decreases, non-foliage surfaces (e.g. soil, branches, dead plant material) become exposed thereby changing the 
reflectance spectrum. The energy reflected is therefore different both qualitatively and quantitatively from that of 
healthy growing vegetation even though the reflectance characteristics of individual leaves may not have changed 
greatly. (Knipling, 1970 (2)). The ability to distinguish healthy and stressed vegetation using this method is 
limited as the differences occur at fairly advanced stages of plant stress. 
However, when changes in the leaf reflectance occur in addition to reductions in leaf area, it is not necessarily the 
case that the canopy is stressed For example, leav es of some species of plants tend to changes their orientation to 
a more vertical profile and curl up, so as to avoid excessive water loss during the hours of strong sunshine. 
Whether it is stress or not the effect here is essentially the same, for the drooping of the leaves reduces the amount 
of foliage and increases the amount of background surfaces exposed to the sensor. (Knipling, 1970 (2)). 
Instrumentation should therefore consider the time-of-day and temperature when attempting to identify stress. 
3.4. The Red Edge 
Using course resolution, broad band spectral data, has shown to be of limited value in characterising plant 
properties. (Boochs et al., 1990 (54). It is anticipated that it will be necessary to use finer resolution 
instrumentation at particular important wavelength bands. One important band occurs around the 700nm-750nm 
region and is known as the 'red-edge'. This occurs because the leaf reflectance changes from very low 1 in the 
chlorophyll red absorption band to very high in the near-infrared (because of internal scattering within the leaf). 
(Horler et al., 1980 (4)). A change in chlorophyll concentration causes the position of the red edge to move within 
the above range. The red edge is a unique feature of live vegetation due to the presence of chlorophyll. 
Leaf spectra of many crops have been recorded during their life cycle. The red edge was shown to shift towards 
longer wavelengths until the onset of senescence and then to shift back towards shorter wavelengths. This is 
consistent with leaves acquiring greater concentrations of chlorophyll in the life cycle and then thing off. It is 
therefore suggested that the position of this edge is a good measure of the amount of chlorophyll present. (Collins. 
1978 (45), Gates et al., 1965 (3)). Horler et al., 1983 (44) and Chappelle et al., 1991 (46) studied the red edge of 
various plant leaves under stressed and non-stressed conditions. The first derivative spectra were used to 
determine the wavelength shift of the red edge and it was found that the first derivative maxima of the red edge of 
vegetation reflectance spectra was a function of chlorophyll concentration. 
However, the effect of a second pigment on the relationship between the red edge and chlorophyll concentration is 
dependent upon the concentration of the second pigment. (Curran et al.. 1991 (47)). The red edge sensitivity to 
the water content of a leaf seems to be species dependent and the position of the red edge either shifts to longer or 
shorter wavelengths depending on how large the change in water content of the leaf is, (Horler et al.. 1983 (44)) 
and the red edge does not seem to be affected by leaf surface moisture, (Horler et al., 1983 (44)). It is unclear as to 
whether the red edge is affected by variations in ground area coverage of a canopy . Horler et al., 1983 (44) say 
that the red edge is not affected but Chappelle et al., 1991 (46) found that there was a slight change in the position 
of the red edge with increasing canopy cover.