The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B7. Beijing 2008
232
Park, Southern Taiwan. The annual mean temperature at the
study area was 22.9°C and annual rainfall was 2,592.4 mm
during 1997-2005. In Nanjenshan Nature Reserve, forest
vegetation distribution is influenced by three major terrains;
windward, valley and leeward (Hsieh and Hsieh, 1990). In this
area the winter precipitation and the intensity of northeast
winds were found to correlate with the differentiation of forest
types. Those in the northeast district are evergreen since the
wind is rather moist. The thorny scrubs and deciduous scrubs
appear in the southwest district due to severe dry winds. The
high species diversity of Renting National Park is largely
attributed to the heterogeneous environment (Su and Su, 1988).
2.2 Plant Materials
Measurement of leaf spectral reflectance was obtained by
randomly harvesting leaves from 4 species from the three types
of terrain. From each terrain type, 1 sampling site from the
leeward area and 2 sampling sites from the windward and
valley areas, respectively, were selected. Fifty sample leaves
taken from the top of the canopies of each species and was
carried out in late April of 2005. Leaf samples were stored in
plastic bags and kept cool for further analysis.
2.3 Measurement of Leaf Chlorophyll Content and
Spectral Reflectance
In this research, the biochemical analysis of chlorophyll
followed the method of Yang et al. (1998). Leaf spectra were
obtained from all the sampling leaves of 4 species which are in
same ages and same size of the same tree species. Leaf
specimens were collected from 3 different terrains pots per
treatment randomly. We obtained a sample size Daphniphyllum
glaucescens of n=35. This was Michelia formosana samples
short of an expected n=34, Illicium dunnianum that is n=32 and
Machilus kusanoi n=26 for the different species and terrains,
respectively. Leaf spectral measurement was conducted using a
GER1500 (Spectra Vista Corporation, NY, USA).
Measurements were taken between 09:30 AM and 14:30 PM.
Conditions varied from cloud-free to overcast skies but care
was taken to avoid measurement when clouds were passing
overhead or darkened.
where the wavelengths for NDVI and SR were 705 and 750 nm,
respectively, and are based on the chlorophyll index developed
by Gitelson and Merzlyak (1994). R 70 5 and R 750 are the leaf
sample spectral reflectance from the GER1500. Based on the
results of Sims and Gamon (2002), they modified these two
indices which tend to increase reflectance across the whole
visible spectrum of a wide range of species. They chose R445
as a measure of surface reflectance and indicated that R is a
445
good reference for all but the lowest chlorophyll content leaves.
The modified indices of NDVI and SR are as follows (Eqs. 3
and 4):
mMX,. =—
705 R 750 +R 705 -2R U5
(3)
^•750 ^445
D _ D
iV 705 - n ’445
(4)
The derivative analysis of spectra reflectance was used
primarily to locate the position and height of the inflection point
of the red edge. The first derivative was calculated using a first-
difference transformation of the reflectance spectrum obtained
from the polynomial fit. In this case, the red-edge peak in the
derivative spectra was composed of a peak maximum usually
between 680 and 750 nm and calculated by the Peakfit curves
statistical software (Version 4.12, Systat Software Inc. San Jose,
USA). The first derivative was calculated using a first-
difference transformation of the reflectance spectrum (Dawson
and Curran, 1998) as follows (Eqs. 5):
(Rj.(j+1) R ÀU) )/M (5)
where FDR is the first derivative reflectance at a wavelength i,
midpoint between wavebands j and j+1, RA,(j) is the reflectance
at the j waveband, RA.(j+l) is the reflectance at the j+1
waveband, and AX is the difference in wavelengths between j
and j+1.
3. RESULT AND DISCUSSION
2.4 Data Analysis
All statistical analyses were conducted using the STATISTICA
statistical software (Version 6.1, StatSoft Inc. Tulsa, Oklahoma,
USA, 2002). Coefficients of determination (R 2 ) were calculated
for relationships between various chlorophyll content from the
result of biochemical analysis treated as independent variables
and vegetation indices and REP were collected from GERÌ500
treated as dependent variable. To test and verify the relationship
of chlorophyll content between vegetation indices and REP,
regression analyses were used in the first data analysis step. The
vegetation indices NDVI (Eqn. 1) and simple ratio (SR) (Eqn. 2)
were used to calculate the vegetative indices obtained from
spectral reflectance measurements:
SR
70S
_ Riso
Ryos
(1)
(2)
Results of the data collection and analyses of chlorophyll
content and leaf reflectance measurements are described here,
showing the relationships found between derivative reflectance
spectra (REP), vegetation indices and chlorophyll content
measurements.
3.1 Spectral Reflectance of Four Species
In order to estimate chlorophyll content of tree leaves using
spectral reflectance, hyperspectral remote sensing provided new
approach for measuring chlorophyll content of plants because
of its hyperspectral analytical rate, variety of wavelengths,
continuance of wavelengths, and abundance of data. On the
aspect of original spectral reflectance, most studies have proven
that there is some correlation between spectral variants and
pigment content. In addition, some predecessors have
discovered that when plants are in different conditions in terms
of nitrogen supply, nutrition level or hereditary factors, the
leaves’ contents of chlorophyll and carotenoid are different, and
the pigments’ contents of different leaves is also closely related
with the environment in which the plant grows. In Fig. 1 we
found that using the first derivatives spectra curve to find REP