Agriculture Reference
In-Depth Information
final regression analysis. Discard elements which are not relevant for a partic-
ular crop species. Weather elements may be considered as single or composite
index. Different composite indices are described in Chapter 3 (Weather), vol-
ume 1. A most commonly used composite thermal index is degree-days, also
called growing-degree-days (GDD), as considered for whole growing period.
(iv)
Regression approach
For a particular zone, multiple regression analysis is to be performed with the
yield series (after subtracting the trend component from original yield series)
and the selected weather elements (single or composite) where yield being the
dependent or explained variable. Different coefficients are obtained for differ-
ent weather elements, and also an intercept or constant value. These values are
used for yield prediction. The predicted yield rate is the sum of the trend com-
ponent and the component for weather elements. The yield (i.e., total yield) is
obtained by multiplying the yield rate (t/ha) with the cultivated area (ha) for
that zone.
The same procedure is to be applied for each of the zones. Finally, the simulation
sub-models may be compacted in one platform to predict the regional or national
yield.
10.5.3.1 The Typical Form of a Model
The typical form of a weather-based multiple regression model is
Y
i
=
C
1
(
i
)
×
t
+
C
2
(
i
)+
C
3
(
i
)
×
GDD(
i
)+
C
4
(
i
)
×
R
(
i
)+
···
where
Y
i
=
predicted yield rate of zone “
i
”
C
1
(
i
)
=
coefficient of trend for zone “
i
”
C
2
(
i
)
=
intercept of regression curve for zone “
i
”
GDD(
i
)
=
growing degree days for zone “
i
”
C
3
(
i
)
=
Coefficient of GDD for zone “
i
”
C
4
(
i
)
=
Coefficient of rainfall for zone “
i
”
R
(
i
)
=
rainfall (cm) for the zone “
i
”
t
=
(simulation year - base year)
The simulation year is the year for which yield is to be predicted.
10.5.3.2 Basic Assumptions in Regression-Based Prediction Model
The basic assumptions in the model are (i) the effect of weather variables on yield
is linear, and (ii) the technological trend over time is similar.
Search WWH ::
Custom Search