Bidirectional Reflectance Measurements
instrument :
spectral range:
spectral resolution:
viewing angle:
illumination angle:
sample measurement:
reference measurement:
IFOV :
laboratory
illumination source:
Iris Mark IV Spectroradiometer
double beam design for two targets:
sample, reference
0.49 - 2.50 pm,
2nm at 0.49 - 1.06 pm 2xSi
4nm at 1.04 - 1.88 pm 2xPbS
6nm at 1.84 - 2.50 pm cooled 243 K
0°
30° Laboratory measurement
3-5 continuous scans
white standard, Halon G-80
parallel to each sample scan,
before and after sample change
on the sample beam
12 x 4 cm 2
at 140 cm distance
quartz-halogene 1000 W
180 cm above sample
Table 3 Measurement configuration and instrument data for the
simulation of multispectral scanner data acquisition,
nadir looking at noon
The measurement configuration and the
instrument for the simulation of
multispectral scanner data
acquisition, nadir looking at noon,
is described in table 3. The high
resolution spectral measurements were
done in one continous scan. The
spectra of the vegetation sample and
the reference (white standard) were
acquired simultaneously, such that
variation in illumination intensity
and spectrum can be corrected for
automatically. The signal level was
monitored for gain and scan time
selection.
The two channels of data are both
stored and recorded for later
processing as two separate spectra
calibrated for radiance, or as a
single ratio spectrum calibrated in
terms of percent reflectance.
4 . Measurement results
Reflectance of single and up to seven
stacked branches of beech (fagus
silvatica) and spruce (picea abies)
was measured from 0.49 - 2.50 pm
wavelength with the IRIS Mark IV
spectroradiometer. Each sample was
measured several times, 5 times in
1988 and 3 times in 1989. For each
measurement the branches were
rearranged. For evaluation the mean
values are presented. The background
had an uniform reflectance of smaller
than ten percent.
In 1988 the measurements focused on
the spectral signature of beech, for
comparison spruce branches were
measured in parallel. In 1989 the
main objectiv was the measurement of
the spectral signature of branches
from damaged beech and spruce trees.
When healthy beech branches were
stacked, the most obvious change in
reflection occurs in the infrared
region (IR). Reflection increases
with every additional branch and
reaches the maximum with 5-7
branches, see figure 1. Reflection is
relatively unchanched in the visible
(vis) and the water absorption region
at 1900 nm. This is consistent with
the interpretation, that pigments of
the first surface layer determine the
spectral response in this wavelength
region. In the infrared reflection is
influenced by the contribution of
multiple transmitted and scattered
radiation and water and/or carbon
dioxide absorption. For the measuring
configuration, as described before,
optical thickness was reached for 5-7
stacked branches.
Figure 2 displays the reflectance of
five stacked branches of beech with
sun leaves (1), with shadow leaves
(2), with discoloured leaves
(yellowing) (3) and spruce (4).
Differences in species and colour
determine the course, mainly the
level of the near infrared plateau,
besides smaller variations in
reflectance of sun leaves and shadow
leaves of beech. The curves follow
the general shape of spectral
signature of green vegetation. In the
visible the course of discoloured
beech branches differs considerably
from those of green beech. Yellowing
produces a raise of reflectance in
the green - red region, a red shift
of the reflection peak of « 10 nm and
a change of its course at the red
edge, which can be attributed to a
narrowing of the chlorophyll
absorption near 680 nm.
Figure 3 shows these effects clearly.
Most prominent is the double peaked
feature in the division spectrum of
the yellowed beech sample. The first
maximum is caused by the increase and
red shift of the reflectance peak.
The second maximum is situated in the
red edge. It is caused by the earlier
52
