Research on the Estimation Model of Vegetation
Water Content in Halophyte Leaves Based on the
Newly Developed Vegetation Indices
Zhe Li, Fei Zhang, Lihua Chen, Haiwei Zhang, and Hsiang-te-Kung
Abstract
The vegetation water content (
VWC
) quantitative is useful for
monitoring vegetation physiological growth. The relation-
ship between
VWC
and vegetation water indices was ana-
lyzed. The optimal estimation model was established. The
results show that: (1) Absorption bands primarily fell within
380 to 400 nm, 680 to 720 nm, 1420 to 1450 nm, 1900 to
1940 nm, and 2450 to 2500 nm; (2) comparing published
vegetation water indices and developed vegetation indices,
it showed that
DVI
(1712,1382)
,
NDSI
(2201,1870)
and
RSI
(2259,1870)
had a
better correlation with
VWC
than the published vegetation
water; and (3)
NDSI
(2201,1870)
and
RSI
(2259,1870)
performed well in
estimating vegetation water content,
DVI
(1712,1382)
had a rough
estimate of its water content. Moreover, the linear combina-
tion of
DVI
(1712,1382)
,
NDSI
(2201,1870)
and
RSI
(2259,1870)
improved the
estimation of
VWC
. The best vegetation indices for estimating
VWC
were found to be the linear combination of
DVI
(1712,1382)
,
NDSI
(2201,1870)
and
RSI
(2259,1870)
in arid area of northwestern China.
Introduction
Water, which controls photosynthesis, respiration, and
biomass, is an important parameter to measure the physi-
ological conditions and morphology of vegetation (Chuvieco
et al
., 2002). In the unit area of the stem and leaf, the gross
mass of liquid water is defined as vegetation water content
(or
VWC
, for short). Hyperspectral remote sensing, developed
in the 1990s, plays a significant role in many fields such as in
agriculture, making irrigation decision through water stress.
It also has been proposed and applied to evaluate forest fire
disasters and crop drought conditions (Ma
et al
., 2017).
The remote sensing technique has surmounted the short-
comings of traditional detection approaches of
VWC
, it extract-
ed more timely and effective information from leaf in land-
scape scale (Davidson
et al
., 2006). Previous studies of
VWC
have shown that hyperspectral remote sensing technology
analysis, the usage of ground vegetation absorption and reflec-
tion characteristics provide a reliable method for monitoring
vegetation water condition. The vegetation indices, which
change in response to vegetation coverage, soil color, leaf
color, and other factors, can also be used to eliminate the
influence of hyperspectral remote sensing technology, and
reflect the vegetation leaf area index, chlorophyll content and
other parameters directly (Fang and Tian, 1998).
For stacks or individual leaf, when
VWC
change,
reflectance in the
NIR
and
SWIR
will also change due to
absorption, with the region of 970 nm, 1200 nm, 1450 nm,
1930 nm and 2500 nm , which are the center of the 5 leaf
water absorption bands (Zhang and Guo, 2006). This finding
lays a foundation for monitoring and estimation of
VWC
.
A variety of studies have been developed to characterize
VWC
. Thomas
et al
. (1971), using fully saturated leaves at
room temperature, discussed the relationship between
VWC
and spectral reflectance preliminarily. They found that the
reflectivity of 1450 nm and 1930 nm bands were significantly
correlated with the relative water content of leaves. Curran
(1989) found that the absorption peak of spectral reflectance
near 1450 nm only reflected the effects of leaf water
absorption, and it peaks at 970 nm, 1200 nm, and 1900 nm
were influenced by other factors, including some nutrients in
vegetation. Dobrowski
et al
. (2005) reported that the canopy
spectra of 690 nm and 740 nm reflected the water stress state
of vegetation. Carter (1991) found that the sensitivity of leaf
reflectance to water content reached its maximum at 1450 nm,
1950 nm, and 2500 nm.
There is a serious shortage of adequate water in arid and
semi-arid regions. Water shortages have been the basic norm
rather than the exception in these regions. If the amounts
of water lost from plants, eventually up to the threshold
value, it can cause premature senescence and reduce the
photosynthetic leaf area of the plant to a level that cannot
sustain growth. The study area is a typical arid and semi-arid
region. Thus, water stress is one of the most severe abiotic
stresses limiting plant growth and productivity in the Ebinur
Lake watershed (Elsayed
et al
., 2017).
However, the choice of the measurement band with high
water absorption is much more complex. Higher incident
energy and low level interference from atmospheric water
vapor are necessary for the measurement of wavebands; this
research shows that the accuracy of
VWC
estimation in the
SWIR
region, which use longer wavelength, is limited (Claudio
et al
, 2006). Thus, this paper tries to find out the optimal
wavelength for
VWC
estimation in the study area.
The goals of the present study were (1) to examine the
reflectance spectra and the first order differential spectra
Zhe Li and Haiwei Zhang are with Resources and Environment
Department, Xinjiang University, Urumqi 830046, P.R.
China; and the Key Laboratory of Oasis Ecology, Ministry of
Education, Xinjiang University, Urumqi 830046, P.R. China.
Fei Zhang is with Resources and Environment Department,
Xinjiang University, Urumqi 830046, P.R.China; the
Key Laboratory of Oasis Ecology, Ministry of Education,
Xinjiang University, Urumqi 830046, P.R. China; and the
Key Laboratory of Xinjiang Wisdom City and Environment
Modeling, Urumqi 830046, P. R. China (
).
Lihua Chen is with the Area Management Bureau of Ebinur Lake
Wetland Natural Reserve, Xinjiang Bole 833400, P.R. China.
Hsiang-te-Kung is with the Department of Earth Sciences, the
University of Memphis, Memphis, TN 38152.
Photogrammetric Engineering & Remote Sensing
Vol. 84, No. 9, September 2018, pp. 537–547.
0099-1112/18/537–547
© 2018 American Society for Photogrammetry
and Remote Sensing
doi: 10.14358/PERS.84.9.537
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
September 2018
537
