of the different water content under the same vegetation
(
Tamarix, Haloxylon ammodendron, Alhagi pseudalhagi,
Phragmites australis, Nitraria tangutorum, Populus
euphratica
) and to analyze the sensitive characteristic
band of the blade; (2) to compare the correlations between
VWC
and published vegetation water indices and newly
developed vegetation indices; and (3) to establish the best
estimation model to provide scientific basis for investigations
of hyperspectral identification and quantitative information
retrieval.
Materials and Methods
Study Area
The study area is located in the northwestern edge of Jinghe
Basin (44°30
′
–45°09
′
N; 82°36
′
–83°50
′
E). Xinjiang Uygur
autonomous region covers an area of 2,670.85 km
2
; the total
area of the Ebinur Lake Watershed is 50.621 km
2
, which is the
national desert ecological reserve and the biggest salt water
lake in Xinjiang (Figure 1). The land is including mountains,
plains, and water bodies. The major water supply source
of nature reserve is from mountain precipitation and melt
mountain glacier. The climate in the Ebinur lake watershed
is a typical continental arid climate, with hot summers and
cold winters; the precipitation is scarce and the evaporation
is high. This surrounding temperature is from 4.0°C to 8.1°C.
The mean annual precipitation is between 102.6 and 229.4
mm from plain to the mountains (Yu
et al
., 2017). The Ebinur
Lake Wetland Nature Reserve in China is the concentrated
distribution area of inland desert salt species. The region has
385 species, accounting for 64 percent
Halophyte
species,; but
the species diversity are low. The dominant species are
Tama-
rix
,
Haloxylon ammodendron
,
Phragmites australis
,
Nitraria
tangutorum
and
Populus euphratica
(Figure 1).
Vegetation Spectrum Measurement
On May 2016, the researchers investigated the Ebinur Lake
Wetland National Nature Reserve vegetation types, soil tex-
ture, and micro topography. We have 71 samples throughout
basin along the Ebinur Lake, and the sample size per species
are as follow:
Haloxylon ammodendro
n (15),
Tamarix
(14),
Alhagi pseudalhagi
(12),
Phragmites australis
(11),
Nitraria
tangutorum
(10) and
Populus euphratica
(9). These species
were the dominant species in the Ebinur lake watershed.
After collecting vegetation leaves back in the laboratory, using
the portable spectroradiometer (Field Spec Hi Res, Analyti-
cal Spectral Devices) an artificial light source was used to
collect each plant spectral curve. The spectral range of the
instrument is 350 to 2500nm with a spectral resolution of 3.0
nm. Before the measurement, we are using the double layer
of dark shades is ensured the light source only from instru-
ment. With the white board calibration, the instrument is 5
cm above each sample; each sample was measured five times,
and the average value was taken as the final spectral value.
Then, the differential transformation and multivariate statisti-
cal analysis are carried out in SPSS and Origin9.1 software.
Leaf Water Content Measurement
After obtaining vegetation samples, they were sealed in plas-
tic bags in time, kept in an ice chest; then carried them to the
laboratory, which were cut into pieces for
VWC
measurements.
Using the analytical balance to get fresh weight, and then put
samples into an oven dry at 80°C for 24 h to determine the
final dry weight values. The fuel moisture content was calcu-
lated as follows:
FMC
=(
FW
–
DW
) /
FW
(
orDW
)×100% (1)
The fuel moisture content (
FMC
), defined as the ratio be-
tween the quantity of water [fresh weight (
FW
) – dry weight
(
DW
)] and either
FW
or
DW
, where
FW
is the fresh weight mea-
sured in the field, and
DW
is the oven dry weight of the same
sample.
FMC
expresses the amount of water in a leaf relative
to the amount of fresh weight or dry matter and is related to
both leaf water content and leaf dry matter content (Wu e
t al
.,
2009).
Vegetation Water Index Selection
The vegetation water index reflects the change in reflectance
of green vegetation in the range of 900 to 2500 nm, which is
influenced by the liquid moisture content of vegetation. This
index is usually composed of two bands (reference band,
measurement band) in the form of the ratio or the normalized
form. Vegetation water indices developed in recent years are
shown in Table 1.
In vegetation spectral studies, the vegetation index can
be extracted quickly and effectively by using derivative
and three side parameters (Gitelson
et al
., 2002), in order
to improve the utilization ratio of spectral information and
eliminate the background noise. But using these methods to
Figure 1. Map of the study area and distribution points.
538
September 2018
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
