×

You are using an outdated browser that does not fully support the intranda viewer.
As a result, some pages may not be displayed correctly.

We recommend you use one of the following browsers:

Full text

Title
Remote sensing for resources development and environmental management
Author
Damen, M. C. J.

266
% of variance between field of the index values
Table 1. Vegetation .indices.
VEGETATION INDEX
. ACR0N0M0US
1
1
BANDS AND
CALCULATION
l|2 1
RATIO
R
1
1
IR/R
1 1
1 1
1 SQUARE RATIO
1 SOR
1
Vir/r
1 § 1
LOG RATIO
Log R
1
1
Log IR/R
1 1
NORMALIZED DIFFERENCES ND
1
1
(IR-R)/(IR.R).0.5
1 1
iOuj I
| TRANSF. VEGETATION
| INDEX
T.V.I.
1
1
V(IR-R)/(IR+R)+0.5
| GREENNESS
1
1 Greenness
1
1
1
1
-aG-bR+cRIR+dIR
l! 1
PERPENDICULAR VEGET.
INDEX
1 P.V.I.
1
1
1
V(R -R ) 2 . (IR .1R ) 2
S V S V
lisi
DIFFERENCES
G
1
1
IR - R
11 1
| DIFFERENCES-DIFFEREN
| CES
! 00
(IR-RIR) 2 - (R-G) 2
1 s 1
1 s 1
G ■ green; R « red;
RIR - red and i
frar
ed; IR « near infrared
% green vegetation cover plotted against calendar
days .
vegetation index-values ( ) and % of variance against
calendar days ( ).
Table II. Coefficient of correlation between vegeta
tion index values and final grain yield versus diffe
rent phenological stages.
tation indices.
maximum correlation between spectral vegetation
values and final grain yield, belong to the
perpendicular and normalized family indices. From
flowering to the end of mulky-maturity phase, the
maximum correlation is related to vegetation indices
based on the difference concept. Before harvest,
that is at the physiological maturity, all the
indices considered in this study show the same
correlation values. The results obtained point out
that it is necessary to adopt different vegetation
indices at specific phenological stages, if the
maximum correlation values have to be exploited to
forecast wheat yield by means of remote sensing
data. This fact can be explained by observing
Figures 1 and 2. In Fig. 1 three spectral profiles
and the trend of the % of the soil vegetation cover
are plotted against seven phenological stages. As
can be observed, the ratio profile function (R) is
more sensitive to the percentage of soil vegetation
cover than the Greenness and Normalized functions
(Greenness, ND). This fact is more evident in
proximity of the maximum soil vegetation cover. That
corresponds approximately to the heading phase.
Since the soil vegetation cover is correlated with
the leaf area index of the wheat (at least for soil
vegetation cover 85%) and the LAI at the heading is
strongly correlated with the final grain yield (if a
normal climatic and environmental condition is