to relate
given by,
OD
\stants a,
d using
stricts of
over the
n found
1gle (38)
x Narrow
of plant
ip has
erse
(i)
important
] of 5-cm
ready for
tion, the
tion. The
ys Since
in
SI/NPACE
(normal,
ice areas
based on decision rule approach. The inverse model is then
applied for the rice area, which uses the relationship of SAR
backscatter, plant height and days since transplantation (see
Table 1) to calculate the transplantation date.
Backscatter co-eff | Predicted plant Days since
(dB) height (cm) transplantation
-18.5 1:3 1
-8 . 2.8 3
-17 6.0 6
-16 9.3 9
-15 12.8 13
-14 16.4 16
-13 20.3 20
-12 24.4 24
-11 28.9 29
-10 33.8 34
-9 39.4 39
-8 45.7 46
-7 83.7 54
-6 65.5 65
Table 1.Relationship between backscatter co-efficient (dB),
predicted height (cm) and days since transplantation.
E
=
= 5
o
$
$497
$ us.
tt
E |
[:]
£A -20 d T T T T T T T 1
0 10 20 30 40 50 66 70 30
Plant height (cm)
Figure 2. Relationship between backscatter co-efficient (dB)
and plant height (cm) used in the study.
A spatial transplantation date map has been generated. This is
suitably colour coded to depict the transplantation pattern of
the study area. This inverse model can be used only for the rice
area. The model can not be used for high backscatter value,
which is mainly related with the nonrice areas because at that
condition the term under square root becomes negative.
5. RESULTS AND DISCUSSION
In case of South 24 Paragana (West Bengal), SAR data of years
2000 and 2001 have been analysed and it was observed that
there is a year to year change in transplantation pattern in the
area. The transplantation patterns are shown in Figure 3a
(2001) and Figure 3b (2000). It has been observed that, in
2001,the transplantation occurred mainly in the month of June
and in 2000, the peak in transplanation is seen in July. This
year to year change in transplantation pattern is found
associated with the rainfall pattern of the area.
The planting of rice crop is mainly depend upon the onset of
monsoon and to verify this fact, the rainfall pattern of West
Bengal for the year 2001 has been analysed. To initiate
IAPRS & SIS, Vol.34, Part 7, “Resource and Environmental Monitoring", Hyderabad, India,2002
transplantation in case of rainfed rice environment, the
availability of cumulative rainfall should be 200 mm. The
rainfall data was collected from the agricultural department.
Four stations were selected to represent the area. The
cumulative rainfall was calculated using the daily rainfall data
and it has been observed that the cumulative rainfall crossed
200 mm during the range from 12th June to 18th June in all the
four stations, which is favourable for transplantation. Hence, a
peak in transplantation is seen in the month of June in 2001 for
South 24 Paragana (West Bengal) The areas showing the
signature of rice during the first fortnight of June is confirmed
as seedbed and grass, which is seen in very small-scattered
pockets.
In case of Orissa, spatial transplantation date map depicting the
progress of transplantation pattern in Baleswar and Bhadrak
districts is shown in Figure 3c.
The cumulative rainfall pattern in case of Baleswar and
Bhadrak districts of Orissa was analysed and found matched
with the transplantation pattern of the area as predicted by the
model (Figure 3c). In Baleswar, as shown in Table 2, assumed
that the distribution of rainfall to be uniform, the cumulative
rainfall was 223 mm on 21st June and transplantation in the
area started after mid June onwards as predicted by the model.
In case of Bhadrak (Table 2), the cumulative rainfall was 255
mm on 30th June and transplantation started from July onwards
(model prediction). The transplantation patterns in both the
areas were confirmed by the ground truth data.
Days Bhadrak Baleswar
rainfall (mm) | rainfall (mm)
June 1st fortnight 120.6 145.7
June 2nd fortnight 135 154.7
July Ist fortnight 652 102.7
July 2nd fortnight 133.3 138.1
August Ist fortnight 158 219.6
August 2nd fortnight 96.7 128.5
Table 2. Rainfall distribution of Baleswar and Bhadrak districts
of Orissa of the year 1999.
The percentage of cumulative variation in rice area with respect
to transplantation date is shown in Figure 4. It is apparent from
the graph that there is a smooth increase in transplantation from
mid June to last week of July, which covers more than 8096 of
the area in both the states. In case of South 24 Paragana in the
year 2001, has been observed that, the percentage of
transplantation is more in the month of June. The graph shows
no such variation in August indicates negligible transplantation
during that period in all the three cases.
Validation of the model has been carried out using six selected
segments in Orissa. The ground data of both plant height as
well as the transplantation date have been used for validation.
The accuracy of the model was examined by calculating the
RMS between the observed value (GT) and estimated value
(model prediction). In case of transplantation date, the RMS is
found to be 5 and for height the RMS value is found to be 1.
The validation of transplantation date as well as plant height as
predicted by the model is shown in Table 3 and Table 4
respectively.