3 RESULTS 
Some results of the model simulations are presented 
in figure 2 to 4. All parameters needed for the cal 
culations are taken from the literature, but in some 
cases not all optical, geometrical or physilogical pa 
rameters were found so that some estimates had to 
be made. 
In our first approach the influence of the pigment 
concentration on the reflectance spectra was analy 
zed. According to our model the pigments are not 
homogeneously distributed in the leaf but are enclo 
sed in small particles , the chloroplasts (Morel and 
Bricaud 1981). Thus the pigment concentration can be 
calculated by two independent methods. First, assu 
ming a constant number of chloroplasts per unit 
volume in the leaf, the total pigment concentration 
of the leaf is given by multipying this number with 
the internal pigment concentration of the chloro 
plasts. By reducing the internal pigment concentra 
tion of the chloroplasts means that the leaf and the 
chloroplasts become more transparent On the other 
hand, assuming a constant internal pigmentation of 
the chloroplasts the leaf becomes more transparent 
by reducing the number of chloroplasts per unit 
volume. In this case the optical properties of the 
chloroplasts are unchanged. 
Running the stochastic model with both versions for 
calculating the total pigment concentration it was 
shown that the reflectance spectrum in the visible 
part is influenced only slightly at the order of less 
than 0.5% over the whole spectrum. In general, hig 
her local pigment concentrations in the chloroplasts 
induce reduced absorption combined with higher re 
flectance. This so called sieve-effect becomes re 
markable only in the absorption bands of chlorophyll 
a by increasing the linewidth. Additional the sieve- 
effect flattens the absorption spectrum of the leaf 
and thus flattens the reflectance spectrum too. 
Based on this result, the dependence of the reflec 
tance spectra on the pigment concentration is analy 
zed by reducing the number of chloroplasts per unit 
volume only. In figure 2 three reflectance spectra 
are shown demonstrating the influence of pigment 
reduction. Due to the decrease of the the pigment 
concentration a reduced absorption is seen by an 
increase of the reflection assocoated with a blue 
shift of the red edge. The increase of the calculated 
reflectance at e.g. 680 nm is about 6%, when the 
pigment concentration decreases to about 1/16 of the 
initial concentration which was about 3 mg/cm 3 of 
chlorophyll a and b. The shift of the red edge is 
typical nm, when the leaf color changes from fully 
green leaves (3 mg/cm 3 ) to light-green leaves (about 
0.2 mg/cm 3 ). This result is in very good agreement 
with the experimental results of Buschmann and 
Lichtenthaler (1988) for cherry-laurel leaves, having 
comparable pigmentation. 
Also in the green spectral range at about 560 nm a 
quite good agreement between the modele reflec 
tance spectra and the experimental data of Bu 
schmann and Lichtenthaler (1988) is seen. The incre 
ase of the reflectance at 560 nm by decreasing the 
pigment concentration at an equal amount is nearly 
identical for the measured and modeled data. The 
model shows an increase of the reflectance from 9% 
for fully green leaves to 25% for light green leaves 
while the experimental data for cherry-laurel leaves 
increase from 10% to 26%. 
For all calculations a difference between the mo 
delled reflectance spectra and the measured data is 
seen in the near-infrared region. In this spectral 
range the internal cell structure and thus the scat 
tering coefficient determines the reflectance spec 
tra, because pigment and water absorption is negli 
gible. On the other hand the water content of the 
leaf determines the arrangement and the optical 
cross section of the cells indirectly and thus the 
scattering coefficient. In our approach, presented in 
this paper, the scattering coefficient is held con 
stant. leading to constant reflectance in the near-in 
frared region. 
Based on the good agreement of the calculated re 
flectance spectra with the data found in the litera 
ture, the seasonal cycle of the reflectance features 
of oak leaves was calculated. The annual variation of 
different leaf pigments as chlorophyll a and b, lu 
tein, violaxanthin, neoxanthin and anthocyanin is gi 
ven by Sanger (1971) , while the specific spectral ab 
sorption coefficients for these pigments are found in 
Lichtenthaler (1987b) and Thimann and Edmondson 
(1949). 
Figure 2: Reflectance spectra of the model leaf depending 
on the pigment concentration. All other leaf parameter 
are held constant. 
I: maximum concentration c 0 
2: reduced concentration c Q / 4 
3: reduced concentration c Q / 16 
478