ratios and potential scaling issues (Yoshioka et al. 2000). Ratios also fail to
account for canopy-background and mutual canopy shadowing associated with
bidirectional reflectance anisotropies.
1.2.2 Vegetation Index and Vegetation Biophysical Isolines
Key theoretical differences among VIs can be depicted through NIR-Red band
cross plots (Fig. 1.3a). A triangular cloud of points delineated by vegetation and
soil end members characterize pixels of varying vegetation amounts and different
canopy backgrounds representing a range of landscape surface conditions. The
canopy background baseline is located close to the 1:1 line and represents the
boundary condition of ‘zero’ vegetation, sometimes known as the ‘‘soil line’’
(Richardson and Wiegand 1977). The end members consist of dense green vegetation at the highest NIR and lowest Red reflectances, bright canopy backgrounds
(dry soil, cement, or snow) and dark backgrounds (wet soil, organic soil, or
standing water). The dense green apex exhibits the maximum Red-NIR contrast
while the soil baseline shows the least differences between Red and NIR
reflectances.
The points inside the triangular structure are mixed pixels composed of spectral
signals from vegetation and canopy backgrounds. As noted by Graetz (1990), over
70 % of the Earth’s terrestrial surface is classified as open canopies, with mixed
background and vegetation signals. Relative to surfaces devoid of plants, partially
vegetated pixels shift away from the lower baseline toward the apex of maximum
NIR and lowest Red reflectance in a manner dependent upon the optical and
structural properties of the canopy and soil background type (Fig. 1.3a). The
greater the amount of green vegetation present in a pixel, the greater will be its’
Red-NIR contrast and shift from the lower soil line, and there is also a convergence toward the apex of maximum vegetation density. The theoretical basis and
defining characteristic of VIs are in how they model these spectrally mixed pixels,
their boundary conditions, and associated variations in time and space.
The SR and NDVI characterize variations in NIR-Red space with vegetation
index isolines of increasing slopes diverging out from the origin, i.e., isoline slopes
become steeper with increasing vegetation amounts but their NIR intercepts are
constant, and at zero (Fig. 1.3b). Isolines of constant VI values define the manner
in which the VI quantifies the subpixel amounts of vegetation present. The NDVI
approximates the soil baseline with an isoline defined by an NIR-Red slope
between 1.25–1.5, which yields NDVI values of 0.10–0.20 for non-vegetated areas
(see Eq. 1.3).
1 Indices of Vegetation Activity
5
account for canopy-background and mutual canopy shadowing associated with
bidirectional reflectance anisotropies.
1.2.2 Vegetation Index and Vegetation Biophysical Isolines
Key theoretical differences among VIs can be depicted through NIR-Red band
cross plots (Fig. 1.3a). A triangular cloud of points delineated by vegetation and
soil end members characterize pixels of varying vegetation amounts and different
canopy backgrounds representing a range of landscape surface conditions. The
canopy background baseline is located close to the 1:1 line and represents the
boundary condition of ‘zero’ vegetation, sometimes known as the ‘‘soil line’’
(Richardson and Wiegand 1977). The end members consist of dense green vegetation at the highest NIR and lowest Red reflectances, bright canopy backgrounds
(dry soil, cement, or snow) and dark backgrounds (wet soil, organic soil, or
standing water). The dense green apex exhibits the maximum Red-NIR contrast
while the soil baseline shows the least differences between Red and NIR
reflectances.
The points inside the triangular structure are mixed pixels composed of spectral
signals from vegetation and canopy backgrounds. As noted by Graetz (1990), over
70 % of the Earth’s terrestrial surface is classified as open canopies, with mixed
background and vegetation signals. Relative to surfaces devoid of plants, partially
vegetated pixels shift away from the lower baseline toward the apex of maximum
NIR and lowest Red reflectance in a manner dependent upon the optical and
structural properties of the canopy and soil background type (Fig. 1.3a). The
greater the amount of green vegetation present in a pixel, the greater will be its’
Red-NIR contrast and shift from the lower soil line, and there is also a convergence toward the apex of maximum vegetation density. The theoretical basis and
defining characteristic of VIs are in how they model these spectrally mixed pixels,
their boundary conditions, and associated variations in time and space.
The SR and NDVI characterize variations in NIR-Red space with vegetation
index isolines of increasing slopes diverging out from the origin, i.e., isoline slopes
become steeper with increasing vegetation amounts but their NIR intercepts are
constant, and at zero (Fig. 1.3b). Isolines of constant VI values define the manner
in which the VI quantifies the subpixel amounts of vegetation present. The NDVI
approximates the soil baseline with an isoline defined by an NIR-Red slope
between 1.25–1.5, which yields NDVI values of 0.10–0.20 for non-vegetated areas
(see Eq. 1.3).
1 Indices of Vegetation Activity
5