ratios and differences (Gallo and Eidenshink 1988), weighted differences and
normalized differences (Tucker 1979), linear spectral band combinations, anglebased VIs (Jiang et al. 2006a), and optimized spectral band combinations (Gobron
et al. 2000; Verstraete and Pinty 1996; Huete 1988; Qi et al. 1994). All VIs relate
positively with plant canopy biophysical properties, and to a certain extent, some
are functionally equivalent. Despite this, there are important differences in how
they depict vegetation foliage and multiple VIs offer a more complete characterization of vegetation canopies. For example, in Fig. 1.2 there is a twofold difference in enhanced vegetation index (EVI) values (0.30–0.65) across forest
canopy types where there is a much smaller range in normalized difference vegetation index (NDVI) values (0.70–0.90).
The Simple Ratio (SR), computed as a ratio of NIR and Red reflectances, is the
most fundamental greenness measure of vegetation in a pixel (Rouse et al. 1973),
SR ¼ q NIR = q red
ð1:1Þ
where q is reflectance. Similarly a vegetation water index (WI) is constructed as
the ratio of NIR with a water-absorbing SWIR band, as in the Moisture Stress
Index (MSI) (Hunt and Rock 1989),
MSI ¼ q NIR = q SWIR
ð1:2Þ
Normalized versions of these ratios constrain values between -1 and +1, e.g.,
the NDVI (Tucker 1979), which is functionally equivalent to SR,
NDVI ¼ SR À 1
ð
Þ= SR þ 1
ð
Þ ¼ q NIR À q red
ð
Þ= q NIR þ q red
ð
Þ
ð 1:3Þ
Fig. 1.1 Field spectral signatures of green leaf, senesced grass, woody bark, and dry and wet
soil. As measured in situ with an Analytical Spectral Devices-Pro, ASD spectroradiometer
1 Indices of Vegetation Activity
3
normalized differences (Tucker 1979), linear spectral band combinations, anglebased VIs (Jiang et al. 2006a), and optimized spectral band combinations (Gobron
et al. 2000; Verstraete and Pinty 1996; Huete 1988; Qi et al. 1994). All VIs relate
positively with plant canopy biophysical properties, and to a certain extent, some
are functionally equivalent. Despite this, there are important differences in how
they depict vegetation foliage and multiple VIs offer a more complete characterization of vegetation canopies. For example, in Fig. 1.2 there is a twofold difference in enhanced vegetation index (EVI) values (0.30–0.65) across forest
canopy types where there is a much smaller range in normalized difference vegetation index (NDVI) values (0.70–0.90).
The Simple Ratio (SR), computed as a ratio of NIR and Red reflectances, is the
most fundamental greenness measure of vegetation in a pixel (Rouse et al. 1973),
SR ¼ q NIR = q red
ð1:1Þ
where q is reflectance. Similarly a vegetation water index (WI) is constructed as
the ratio of NIR with a water-absorbing SWIR band, as in the Moisture Stress
Index (MSI) (Hunt and Rock 1989),
MSI ¼ q NIR = q SWIR
ð1:2Þ
Normalized versions of these ratios constrain values between -1 and +1, e.g.,
the NDVI (Tucker 1979), which is functionally equivalent to SR,
NDVI ¼ SR À 1
ð
Þ= SR þ 1
ð
Þ ¼ q NIR À q red
ð
Þ= q NIR þ q red
ð
Þ
ð 1:3Þ
Fig. 1.1 Field spectral signatures of green leaf, senesced grass, woody bark, and dry and wet
soil. As measured in situ with an Analytical Spectral Devices-Pro, ASD spectroradiometer
1 Indices of Vegetation Activity
3