Você está na página 1de 8

4th International Conference on Water & Flood Management (ICWFM-2013)

ESTIMATION OF THE CHANGES OF WETLANDS IN THE


NORTHWEST REGION OF BANGLADESH USING LANDSAT IMAGES
Ahsan Azhar Shopan1*, A.K.M. Saiful Islam2 , Nepal Chandra Dey3 and Sujit Kumar Bala4
1

Research and Evaluation Division, Bangladesh Rural Advancement Committee (BRAC),


75 Mohakhali, Dhaka-1212, Bangladesh, e-mail: ahsan.shopan@gmail.com
2
Institute of Water and Flood Management, Bangladesh University of Engineering and
Technology, Dhaka-1000, Bangladesh, e-mail: akmsaifulislam@iwfm.buet.ac.bd
3
Research and Evaluation Division, Bangladesh Rural Advancement Committee (BRAC),
75 Mohakhali, Dhaka-1212, Bangladesh, e-mail: nepal_dey@yahoo.com
4
Institute of Water and Flood Management, Bangladesh University of Engineering and
Technology, Dhaka-1000, Bangladesh, e-mail: bala@iwfm.buet.ac.bd
ABSTRACT
Bangladesh has a high proportion of wetland areas, which of late have been declining. In the
Ganges-Brahmaputra floodplain alone, approximately 2.1 million ha of wetland have been lost to
flood control, drainage and irrigation development. This study estimates the changes of wetlands
in the Northwest Bangladesh by analyzing Landsat satellite images using remote sensing and GIS
tool, ILWIS. For this purpose, three Landsat scenes of 1989, 2000 and 2010 are used. It is found
that, total wetland area in the Northwest region of Bangladesh during the dry period in 1989,
2000 and 2010 were 1208.72, 903.54 and 867.18 sq. km respectively. This indicates a decrease of
25.25% wetland areas from 1989 to 2000 and a decrease of 4.02% wetland areas from 2000 to
2010. The changes of wetland areas during the dry periods of different years in the selected
districts and upazilas of the Northwest Bangladesh are also estimated. Wetlands in most of the
selected districts and upazilas have been decreased during the period of 1989-2010.
Keywords: Wetlands, wetland changes, landsat, NDVI, remote sensing
1. INTRODUCTION
Bangladesh has a high proportion of wetland areas, which have been declining in recent years. In
the Ganges-Brahmaputra floodplain alone, approximately 2.1 million ha of wetlands have been
lost due to flood control, drainage and irrigation development. Wetlands are being continuously
lost or degraded primarily because of various recent developments, such as shrimp culture, which
reflect a lack of community awareness of wetland functions and values. The exploitation of the
Haor wetland ecosystem began due to ever-expanding agrarian settlements. The Beels are being
drained, and embankments built to save crops from flash floods. Swamp forests that were once
extensively distributed, are now on the verge of extermination (UNEP, 2001).
Islam et al. (2010) identified the wetlands in the Dhaka city, emphasizing on the extent of their
changes and subsequent impacts on physical environment using topographical map (Toposheet of
1960) and satellite images (Landsat TM, 1988 and ALOS VNIR, 2008). The study observed that
in the Dhaka city water bodies and lowlands decreased by 32.57% and 52.58%, respectively
during 1960 and 2008.

521

4th International Conference on Water & Flood Management (ICWFM-2013)


Sultana et al. (2009) studied pre- and post-urban changes of water bodies in the Dhaka city using
aerial photographs and SPOT images. In 1968, the total area of marshy and peaty inundated lowlying areas was 133 km2, which was depicted to be 67 km2 in the year 2001. The total area of
inland lakes in 1968 was 5.1 km2 which became 1.8 km2 in the year 2001. The urban
infrastructures filled and/or compartmentalized the water bodies, causing water loggings problem
during wet-season in various part of the city.
Mahmud et al. (2011) evaluated wetland changes in Dhaka Metropolitan Area between 1978 and
2009. Spatial and temporal dynamics of wetland changes were quantified using four Landsat
images, a supervised classification algorithm and the post-classification change detection
technique in GIS environment. Accuracy of the Landsat-derived wetland maps ranged from 87%
to 92.5%. The analysis revealed that area of wetland and Rivers & Khals in Dhaka city decreased
significantly over the last 30 years by 76.67% and 18.72% respectively. This changing trend of
wetlands makes the drainage system of Dhaka City vulnerable, creating water logging problems
and their consequences. Land filling and encroachment were recognized to be the main reasons for
shrinking of the wetlands in the city. However, no study has been found which uses remote
sensing for detecting changes of the wetlands of the Northwest region of Bangladesh.
The northwest region of Bangladesh is particularly vulnerable to droughts. Apart from loss to
agriculture, droughts have significant effect on land degradation, livestock population,
employment and health. Fresh water resources in drought-prone areas such as the northwest
Bangladesh are already declining due to over exploitation to support irrigation in the dry months.
It has been predicted that by 2018, the demand for irrigation may reach 58.6 percent of the total
supply. (Ramamasy and Baas, 2007). The reduction of wetlands in the Northwest region of
Bangladesh will further enhance drought and environmental degradations in this area.
In this context, this research work has been carried out to determine the spatial coverage area of
wetlands in the northwest region of Bangladesh during the dry period of 1989, 2000 and 2010
using Landsat images and estimate the changes in wetland area over the period of last 21 years.
Changes of wetlands during the dry period in selected districts and upazilas of the Northwest
region of Bangladesh will also be estimated.
2. METHODOLOGY
2.1 Data
In order to compare the water bodies of Northwest Bangladesh, three Landsat scenes of 1989,
2000 and 2010 are used. Spatial resolution of the satellite images is 30m and they have been
downloaded from www.earthexplorer.usgs.gov.
Table 1. Properties of downloaded satellite images
Image No.
1
2
3

Acquisition Date
January 19, 1989
February 19, 2000
February 6, 2010

Satellite Sensor
Landsat 4 TM
Landsat 7 ETM+
Landsat 5 TM

These images are taken at Path of 138 with Row of 42 and Path of 138 with Row of 43 which
cover almost the entire Northwest Bangladesh (about 90% of the total area). Images represent dry
522

4th International Conference on Water & Flood Management (ICWFM-2013)


season of Bangladesh as they have been captured in Late January to Late February on different
images. It is assumed that temporal changes of water bodies remain insignificant over this period.
Properties of the images are presented in Table 1.
2.2 Image Processing
Integrated Land and Water Information System (ILWIS, 2012) has been used to process the
Landsat images and conduct spatial analysis. Pre-processing of satellite data includes many
intermediate steps such as image importing, subsetting, conversion to radiance, conversion to
reflectance etc. In order to determine the area of wetlands in the Northwest region of Bangladesh,
the NDVI maps for the selected area are prepared from satellite images. Then all the pixels with
positive NDVI values are removed. It is known that negative values of NDVI correspond to water.
So, the remaining pixels with negative NDVI indicate wetlands or water bodies. Total pixel area,
corresponding to the total wetland area is obtained from map histogram in square meter. This area
is then converted in to square kilometers. Total wetland area determined from each of the satellite
images is then used to estimate the changes in wetland areas of the Northwest region of
Bangladesh.
2.3 Normalized Difference Vegetation Index (NDVI)
The Normalized Difference Vegetation Index (NDVI) is a measure of the amount and vigor of
vegetation at the surface. Bala and Islam (2008) have shown the effectiveness of using NDVI data
for prediction of yield of potato in Bangladesh. The reason of NDVI related to wetland is that
water reflects more in the visible red part than in the near infrared part of the spectrum. The index
is defined by Equation 1.

Figure 1: NDVI map

523

4th International Conference on Water & Flood Management (ICWFM-2013)


(1)
It is clear from the above equation that water bodies or wetlands would give negative NDVI values
as for water, RedNIR. Figure 1 shows the NDVI map of the Northwest region of Bangladesh.
2.4 Wetland Area
The following rule has been set to determine the area of wetland from NDVI map: Wetland =
NDVI0. Figure 2 shows a map of wetland area in the Northwest region of Bangladesh on 6
February 2010.

Figure 2: Map of wetland area in the northwest region of Bangladesh on 6 February 2010
3. RESULTS AND DISCUSSION
Table 2 shows the total wetland area during the dry period at different years in the Northwest
region of Bangladesh. Table 3 shows the change in wetland areas in the Northwest region of
Bangladesh over the years. Figure 3 shows the decreasing trend of wetland areas.
Table 2: Total wetland area in northwest Bangladesh during the dry period
Year
1989
2000
2010

Wetland Area (km2)


1208.72
903.54
867.18

524

4th International Conference on Water & Flood Management (ICWFM-2013)


Table 3: Changes in wetland areas in the northwest region of Bangladesh during the dry period
Time Period
To
Year
Interval
1989 2000
11
2000 2010
10

From

Change in
Wetland
Area (km2)
-305.18
-36.36

Percentage
Change (%)
-25.25
-4.02

Change in
Wetland Area
per Year (km2)
-27.74
-3.64

Percentage
Change per
Year (%)
-2.30
-0.40

Figure 3: Change in wetland area in northwest Bangladesh during the dry period
It is found that, about the one-third of the total wetlands in the Northwest Bangladesh have been
lost over the period of 21 years from 1989 to 2000. However, the rate of decreasing is much lower
during 2000-2010 than that of during 1989-2000.

1989

2000

2010

Figure 4: Chronological maps of wetland Area of northwest Bangladesh during the dry Periods in
the years 1989, 2000 and 2010

525

4th International Conference on Water & Flood Management (ICWFM-2013)


Figure 4 shows chronological maps of wetland area of the Northwest Bangladesh during the dry
periods in the years 1989, 2000 and 2010.Table 4 shows the total wetland areas in the selected
districts of the Northwest Bangladesh during the dry periods in the years 1989, 2000 and 2010.
Rajshahi, Pabna, Bogra, Rangpur and Dinajpur districts are located in the Northwest Bangladesh.
Table 5 shows the changes of wetland areas in the selected districts over the period.
Table 4: Total wetland areas in the selected districts during the dry period
Wetland Area (km2)
Year
District
Rajshahi
Pabna
Bogra
Rangpur
1989
79.89
123.02
125.13
24.04
2000
67.58
98.31
104.82
11.21
2010
52.59
76.85
94.57
12.78
Note: Dinajpur district not fully covered by satellite imagery (about 95% covered)

Dinajpur
15.54
12.52
17.09

Table 5: Changes of wetland areas in the selected districts during the dry period

District
Rajshahi
Pabna
Bogra
Rangpur
Dinajpur

From
1989
2000
1989
2000
1989
2000
1989
2000
1989
2000

Time Period
To
Year
Interval
2000
2010
2000
2010
2000
2010
2000
2010
2000
2010

11
10
11
10
11
10
11
10
11
10

Change in
Wetland
Area (km2)

Percentage
Change
(%)

-12.31
-14.99
-24.71
-21.46
-20.31
-10.25
-12.83
1.57
-3.02
4.57

-15.41
-22.18
-20.09
-21.83
-16.23
-9.78
-53.37
14.00
-19.43
36.50

Change in
Wetland
Area per
Year (km2)
-1.12
-1.50
-2.25
-2.15
-1.85
-1.03
-1.17
0.16
-0.27
0.46

Percentage
Change
per Year
(%)
-1.40
-2.22
-1.83
-2.18
-1.48
-0.98
-4.85
1.40
-1.77
3.65

From tables above, it is found that, in all of the selected districts except Dinajpur, wetlands have
decreased since 1989. However, in Rangpur district wetlands have increased nominally during
2000-2010 after it lost more than half of its wetland areas during 1989-2000. Among the districts
selected from Rajshahi division, Pabna lost the highest percentage of wetlands since 1989.
Table 6 shows the total wetland area during the dry period in different years in the selected
upazilas of the Northwest of Bangladesh (one upazila from each of the districts previously
selected). The five selected upazilas are: Godagari upazila of Rajshahi district, Chatmohor upazila
of Pabna district, Dhupchanchia upazila of Bogra district, Badarganj upazila of Rangpur district
and Birampur upazila of Dinajpur district.

526

4th International Conference on Water & Flood Management (ICWFM-2013)


Table 6: Total wetland areas in the selected upazilas during the dry period

Year

Godagari
(Rajshahi)
9.84
11.55
3.23

1989
2000
2010

Wetland Area (km2)


Upazila (District)
Chatmohor
Dhupchanchia
Badarganj
(Pabna)
(Bogra)
(Rangpur)
3.22
0.20
1.31
1.91
0.26
0.48
2.54
0.05
1.11

Birampur
(Dinajpur)
0.47
0.51
0.70

Table 7: Changes of wetland areas in the selected upazilas during the dry period

District
Godagari
Chatmohor
Dhupchanchia
Badarganj
Birampur

From
1989
2000
1989
2000
1989
2000
1989
2000
1989
2000

Time Period
To
Year
Interval
2000
2010
2000
2010
2000
2010
2000
2010
2000
2010

11
10
11
10
11
10
11
10
11
10

Change in
Wetland
Area
(km2)
1.71
-8.32
-1.31
0.63
0.06
-0.21
-0.83
0.63
0.04
0.19

Percentage
Change
(%)
17.34
-72.03
-40.68
32.98
30.00
-80.77
-63.36
131.25
8.51
37.25

Change in
Wetland
Area per
Year (km2)
0.155
-0.832
-0.119
0.063
0.005
-0.021
0.083
0.063
0.004
0.019

Percentage
Change
per Year
(%)
1.58
-7.20
-3.70
3.30
2.73
-8.08
-5.76
13.13
0.77
3.73

Among the selected upazilas, except Birampur of Dinajpur district, all other upazilas have lost
significant percentage of wetlands since 1989. On the other hand, wetlands in Birampur upazila
show a continuous increasing trend from 1989 to 2010. Considering only the time period of 20002010, wetlands in Badarganj upazila of Rangpur have increased about twice of its area, which is
the highest during this period. However, Dhupchanchia upazila of Bogra district and Godagari
upazila of Rajshahi district have lost more than 80% and 70% of their wetlands during the same
period.
4. CONCLUSION
Results of the analysis of satellite images make it evident that wetlands in the Northwest region
Bangladesh have a decreasing trend. During 1989 to 2000, about 25% of wetlands have been lost.
However, from 2000 to 2010, about 4% wetlands are converted into other types of land. Sharp
decreasing trends are observed in Rajshahi, Bogura and Pabna district. Total wetland areas in
Rangpur district has been increased slightly from 2000 to 2010 though it is decreasing
approximately about 50% from 1989 to 2000. Wetlands in Dinajpur have moderately decreased
over the last 21 years. For the five selected Upazilas, except Birampur Upazila, the net wetlands
have been lost over the last two decades. Severe loss of wetlands is found in Godagari Upazila of
Rajshahi district. This is affecting the marine resources, agriculture and groundwater level in the
northwest region of Bangladesh.

527

4th International Conference on Water & Flood Management (ICWFM-2013)


REFERENCES
Bala, S.K. and Islam, A.S. (2008). Estimation of potato yield in and around of Munshiganj using
remote sensing NDVI data. Institute of Water and Flood Management (IWFM), BUET,
Dhaka-1000, Bangladesh.
ILWIS (2012), Integrated Land and Water Information System, [online], available from
http://52north.org (accessed on 25 August 2012).
Islam, M.S., Rahman, M.R., Shahabuddin, A.K.M. and Ahmed, R. (2011). Changes in Wetlands
in Dhaka City: Trends and Physico-Environmental Consequences. Journal of Life and Earth
Science, University of Rajshahi, Vol. 3, 2011, 781-787.
Mahmud, M.S., Masrur, A., Ishtiaque, A., Haider, F. and Habiba, U. (2011). Remote Sensing and
GIS Based Spatio Temporal Change Analysis of Wetland in Dhaka City, Bangldesh.
Journal of Water Resources and Protection, Vol. 3, 2011, 781-787.
Ramamasy, S. and Baas, S. (2007). Climate variability and change: Adaptation to drought in
Bangladesh. Prepared under the Asian Disaster Preparedness Centre and Food and
Agriculture Organization of the United Nations, Rome, Italy.
Sultana, M.S., Islam, G.M.T. and Islam, Z. (2009). Pre- and Post-Urban Wetland Area in Dhaka
City, Bangladesh: A Remote Sensing and GIS Analysis. Journal of Water Resources and
Protection, Vol. 1, No. 6, 2009, 414- 421.
UNEP (2001). Bangladesh: State of the Environment 2001. United Nations Environment
Program (UNEP), Thailand.

528

Você também pode gostar