n be
ative
and
der
BY
| LU
f the
[CA)
[ the
itical
have
10WN
rk is
Yodel
ying
ntal,
wing
es in
. For
ation
duce
work
linity
ation
more
3orai
flow
linity
0 an
rface
lated
IAPRS & SIS, Vol.34, Part 7, “Resource and Environmental Monitoring”, Hyderabad, India,2002
surface water salinity data at designated points were used to
generate iso-haline surfaces within a GIS environment.
——
rie dei Hol
Water Resource System | |
Management 3cenafós | | Management Strategies
Dry | Land Aras | Adocated Land Argas for |
Semson — by Salinity —» Crop aed Snnmp ^
Salinity | ; Class Production by Salinity Class, |
i d
i Land Une Outputs. |
| Wild Harvie 4. And inputs y |
| Capium Msherios | Salinity Class |
| . Bagda ry
| - Goldsty |
Li. |
Y
Sal ma a {
Environments impacts | Le Econcenie Impacts So impacts
À
Figure 3. Computational framework for the shrimp-paddy
: system
The base condition in this case is a situation where the dry
season flow of the Gorai is close to zero, which is
representative of the hydrological condition of the river in the
1997-98 hydrological year. The models were also run for its
different flow regimes. The simulated salinity data for the Gorai
flows of 60, 100 and 150 m?/s were used to produce the
corresponding salinity intrusion maps. The different salinity
levels available on the continuous iso-haline surface were
grouped together into salinity ranges showing areas of «5ppt, 5-
10ppt, 10-15ppt, 15-20ppt and >20ppt as shown in Figure 4.
3.3 Land-use classification
In the base flow conditions, the major land use classes are
generated from the classification of a combination of various
satellite imagery. The main sources of data listed in Table 6. A
combination of image processing techniques, extensive field
information, knowledge of hydrology, groundwater condition,
existing agricultural practices, GIS data layers of salinity, land
type, river networks, etc., were used in the classification of the
imagery.
For the dry season landuse, four Landsat TM images covering
the entire study area were used. Using all bands except the TM
thermal band, the permanent water class including rivers and
beels or waterbodies were classified. Unsupervised methods of
classification of the rest of the image (excluding the
Sunderbans mangrove forest) produced 30 classes. The six final
classes identified are as follows: permanent water, Boro rice,
mixed winter crop, bare fallow land, fallow land with
grass/stubble, and aquatic vegetation. The Bagda shrimp farms
were classified from the ERS-2 radar image of 18 May, 1999.
The dry season land use map is shown in Figure 5. For the
monsoon land use including the aman rice crop, ERS-2 radar
image classification was used as dense cloud cover exists
throughout the monsoon. Field information was used to extract
signatures from the multi-temporal radar image set. The images
were classified into permanent water areas, Bagda, Aman and
“other vegetation” class as shown in Figure 6 (EGIS &
SPARRSO, 2001). The rural settlements and bagda farms were
on-screen digitized from IRS panchromatic images. The
settlements were identified by the presence of large trees
surrounding the villages and the high boundaries of the Golda
ponds and their rectangular shapes were used for the visual
interpretation of the Golda ponds.
instant Flowe Um sin Cua gi | bra ILLE
Yan Floss Wo omowe Caress Bhrem 1500 rm wer
<5 5-10 W10-15 15-20 #>20
Salinity (ppt)
Figure 4. Surface water salinity at different flows of the
Gorai River
Classification Source Image Date of Source Image
Dry season land use Landsat -5 TM February, 1999
Bagda shrimp farms ERS -2 SAR May, 1999
Monsoon land use ERS 2 SAR May and August, 1999
Rural settlement and IRS-1D Pan January to March, 1998
Golda shrimp farms
Table 6. Image dataset used for land use mapping
3.4 Cropping pattern map
Both the dry and wet season landuse-landcover raster layers
were combined to generate a cropping pattern layer of the study
area on the basis of ‘expert’ and field knowledge. The cropping
pattern was simplified in order to show the following major
land use classes relevant for this study. The “major production
regimes” that are identified through this process are:
(1)Bagda/Aman, (2)Bagda/Bagda, (3)Aman, (4)Golda/Boro,
(5)Aman/Boro. “Other land use” are termed as follows: Mixed
Crops, Other Vegetation, Settlements, Mangroves, Beels/small
rivers.
3.5 Other models and sub-models
Land-use allocation sub-model: For each flow condition
under consideration, the model shows the distribution of the
land use classes among the various salinity regimes and allows
for land use allocation based on the strategies. This is done by a
GIS overlay of the salinity map and the land use map. With
knowledge of this distribution, areas under each salinity class
may be allocated to a particular land use. Allocation is done
only in the areas under the “major crop regimes”. The areas
under “other land use”, as determined above, are not variable.
The allocation schemes are determined by a land use strategy
under consideration during a model run.
Economic assessment model: The model calculates the return
from different production regimes including paddy, shrimp and
fish under different salinity classes. Production of crops varies
according to salinity. The average yield for different crops are
adjusted for different salinity classes. The economic return
557
