Each cg 
| that has 
leaf area 
S obtained 
IN ON each 
measure. 
sthod and 
vided into 
point of a 
d direction 
method is 
solution is 
a of each 
rity image 
y leaf area 
ection and 
oil is given 
this cell 
ard in this 
factor, 1 
| Sy ae 
n the cel 
ndicular to 
'S the cell, 
ributing is 
)sorbed is 
smitting is 
of the cell 
is cell, the 
lent point 
Lamberlian 
e specular 
reflectance factor caused by the ponding water 
is 2.1% here. 
3. REFLECTANCE MEASUREMENT 
31 Measurement of canopy bidirectional 
reflectance factors 
Rice crop (Oliza Sativa, variety: koshihikari) is 
cultivated at an experimental field of University of 
Tokyo in 1992. Dense plot (15cm X 15 cm) 
and sparse plot (30 cm X 30 cm) are set. 
Reflectance measurement is carried out a 
Thus, NIR is suited for deciding canopy 
structure from reflectance characteristics, because 
variations of reflectance and transmittance factors 
among individual leaves are small. The visible region is 
suited for states of individual leaves from reflectance 
characteristics.Reflectance factor p and transmittance 
factor v of ten samples of rice individual leaves are 
measured nondestructively in 1995. Incident angle is O 
deg. and reflection angle is 45 deg.. 
For the result, in the NIR region (850nm - 
  
week interval under artificial light (450W 
  
lump) just after sun set. Reflectance 
factors are measured with 10 deg. FOV. 
Data is obtained three times per one object 
  
  
and averaged. The device for reflectance 
measurement is shown in Fig. 2 shows . 
Incident angle to the canopy is 15 deg.. 
View azimuth angle is O deg. (the opposite 
of the light) and view zenith angles are 0, 
45. and 60 deg.. Soil reflectance (8 cm 
water ponding) is also measured. 
  
  
  
  
  
  
  
3.2 Measurement of individual leaf © 
reflectance factor 
Variation characteristics of spectral reflectance 
and transmittance of rice individual leaves, which are 
seldom investigated in canopy reflectance modeling are 
analyzed. Difference of leaf spectral characteristics 
due to the difference of leaf situations and fertilizer 
levels are inquired. And these characteristics in an 
individual leaf are also inquired. 
Rice crop is cultivated in a pot in a green house. 
The fertilizer levels are high, middle, and no fertilizer. 3 
pots are set in each fertilizer level and data is arranged 
by randomized block design, considering other 
conditions. Reflectance and transmittance factors are 
nondestructively measured in  280-2500nm. The 
incident angle is 45 deg. And reflection angle is O deg. 
Analysis of variance in two-factor factorial 
experiment and inspection using least significant 
difference are carried out. The factors of this factorial 
experiment are leaf situation on a stem and fertilizer 
level. In NIR, there are no significant differences of 
reflectance and transmittance factor due to the leaf 
Situations and fertilizer levels among the fertilizer 
applied plots. In the visible region, there are significant 
differences of reflectance factor due to the leaf 
Situations. The coefficients of variations of all the 
reflectance and transmittance factors (number of 
Samples: 41) are as small as 5% in NIR (Fig. 3). 
Reflectance and transmittance factors in a individual 
leaf are in the corresponding ranges of the coefficients 
of variation in NIR. On the other hand, in the visible 
region, variation of the factors are rather large. 
Pes 
oF 
Sparse plot Dense plot 
Fig. 2. Canopy reflectance measurement 
transmittance factor (%) Reflectance factor (%) 
e ) 
391 
  
© 
(=>) 
A 
e 
("NS 
e 
  
  
> 
e 
N 
e 
  
  
  
  
n° 
00 600 1000 1400 1800. 
Wavelength (nm) 
Fig. 3. Spectral characteristics of rice leaves 
(average S.D., n241) 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B7. Vienna 1996