Transcript Document

Soil-Adjusted Vegetation Index
A transformation technique to minimize soil brightness
from spectral vegetation indices involving red and nearinfrared (NIR) wave lengths.
The transformation involves a shifting of the origin of
reflectance spectra plotted in NIR-red wavelength space to
account for first-order soil-vegetation interactions and
differential red and NIR flux extinction through vegetated
canopies.
Near InfraRed (NIR)
Multispectral
Arizona Fire - USA
21-JUN-2003
In general, most vegetation indices rely in the
existence of a “soil line” in red and NIR
wavelength space, i.e., there is a principal axis
of soil spectral variation extending outward
from the origin with increasing brightness.
Since most of the soil spectra fall on or close o
the soil line, and since the intercept of such a
line is close to the origin, RVI and NDVI values
of bare soils (ratios) will be nearly identical for
a wide range in soil conditions.
Source: Huete, A.R. 1988
Figure 2
Source: Huete, A.R. 1988
NIR  l2
red  l1
NIR  l2   red  l1 
NIR  l2   red  l1 
Since the soil line has slope close to 1, the adjustment factors, l1
and l2, would be nearly equivalent. Shifting the red and NIR data
equally (l1=l2) and utilizing the NDVI format
NIR  red
NIR  red  L
Where L=l1+l2=2l. Thus a soil adjustment index(SAVI) would only
involve an addition of a constant, L, to the denominator of the
NDVI equation.
However, in order to maintain the bounded cinditions ot the NDVI
equation (NDVI can vary from –1 to +1), a multiplication factor
(1+L) is needed in eq. 3
NIR  red
SAVI 
 1  L 
NIR  red  L
Figure 3
Source: Huete, A.R. 1988
Figure 4
Source: Huete, A.R. 1988
Figure 5
Source: Huete, A.R. 1988
MODIFIED-SAVI

2 NIR  1  2 NIR  1
MSAVI 
2
2
 
 8NIR  red  0.5
Reference
• Huete, A.R. 1988. “A soil-Adjusted Vegetation Index (SAVI)”,
Remote Sensing of Environment, 25:295-309
•Huete, A.R., Lui, H.Q. 1994. “An Error and Sensitivity
Analysis of the Atmospheric and Soil- Correctin Variants of the
NDVI for he MODIS-EOS.” IEEE Transactions on Geoscience
and Remote Sensing, 32(4), 897-905