Remote Sensing of Leaf, Canopy, and Vegetation Water Contents for Satellite Environmental Data Records - Satellite-based Applications on Climate Change - page 329

Fig. 20.2 Spectral reflectances for fresh and dehydrated leaves of corn ( Zea mays ). Along the top
are the wavelength positions of the Moderate Resolution Imaging Spectroradiometer ( MODIS )
land/cloud/aerosol bands (250 and 500-m pixels) and along the bottom are the wavelength
positions of the Thematic Mapper ( TM ) and Enhanced Thematic Mapper Plus ( ETM +) bands.
MODIS ocean color and water vapor bands (1,000-m pixels) are not shown

Many vegetation types respond to changes in the season by growth of new leaves
and senescence of old leaves. Different vegetation types have different seasonality
defined by either temperature or precipitation (Tucker et al. 1985 ). By analysis of
multi-temporal NDVI, phenological events such as the start of spring may be
determined (Schwarz et al. 2002 ). As a response to stress, many vegetation types
lose some or all of their leaves, so differences in NDVI compared to the annual
maximum and minimum NDVI become powerful indicators of vegetation stress
(Kogan et al. 2003 ; Brown et al. 2008 ). Changes in LAI will proportionately affect
CWC (Table 20.1 ), so annual changes in remotely sensed CWC should be similar to
the annual changes in NDVI or other red/near-infrared index.
If changes in remotely sensed CWC simply mirrored changes in NDVI, there
would be no use for CWC data products. From the previous section, single
measurements of LWC and CWC values cannot indicate physiological measures
of stress (RWC or

Ψ

). However, reductions in LWC and CWC occur with drought
stress before changes in LAI, so a high frequency of CWC observations may yield
more reliable estimates of drought stress than NDVI. Quantifying physiological
water stress accurately over time would then become an important environmental
data record.