Independent University Bangladesh (IUB)

IUB Academic Repository

Dept. of Environmental Science and Management Article

2020-09-20

Predictive assessment on landscape

and coastal erosion of Bangladesh
using geospatial techniques

Islam, Md. Mozahidul

Elsevier

[Link]
Downloaded from IUB Academic Repository
 Remote Sensing Applications: Society and Environment 17 (2020) 100277

Contents lists available at ScienceDirect

Remote Sensing Applications: Society and Environment

journal homepage: [Link]

Predictive assessment on landscape and coastal erosion of Bangladesh using

geospatial techniques
Md. Mozahidul Islam a, *, Md. Saifur Rahman b, Md. Alamgir Kabir c, Md. Nazrul Islam d,
Ruhul Mohaiman Chowdhury e
a
Department of Environmental Management, SESM, Independent University, Bangladesh, Bashundhara R/A, Dhaka, 1229, Bangladesh
b
Department of Geo-Information Science and Earth Observation, Patuakhali Science and Technology University, Dumki, Patuakhali, 8602, Bangladesh
c
Department of Agroforestry, Patuakhali Science and Technology University Dumki, Patuakhali, 8602, Bangladesh
d
Department of Geography and Environment, Jahangirnagar University, Savar, Dhaka, 1342, Bangladesh
e
Sustainable Forests & Livelihoods (SUFAL) Project, Bangladesh Forest Department, Agargaon, Dhaka, 1215, Bangladesh

A R T I C L E I N F O A B S T R A C T

Keywords: Coastal erosion, land use and land cover (LULC) changes analysis using remote sensing is a dynamic, relatively
LULC low cost based precise method using now a day. Coastal districts of Bangladesh occupied by naturally grown
Sub-district mangrove forest which are susceptible to rapid land cover (LC) changes and natural erosion. Barguna and
Accretion-erosion
Patuakhali districts of Bangladesh deserve special attention for conserving coastal mangrove forest named
Mangrove forest
Conversion
Tengragiri Wildlife Sanctuary and variety of human forces income. The core objective of this research is to
analyze the LULC change along with coastal erosion analysis from 2000 to 2017. Combination of four years
Landsat satellite image analysis, primary field data, geo-tag photography, secondary information, utilization of
forest carbon inventory 2015 data, and semi-structured questionnaire are the key approaches adopted in the
study. K-means cluster based unsupervised and maximum likelihood supervised classification by using ERDAS
Imagine 2014 found the total study area is 33,361 ha. Random sampling (40 points/class) based accuracy
assessment and verification by google earth pro 7.1 found overall accuracy 88.15% and Kappa coefficient is
0.867. Python coding program and overlay operation tested for conversion analysis any found weighted overlay
provide best results. An intensive RS analysis of 33,564 ha mangrove forest and community landscapes generated
six (6) distinct land cover class and sub-classes, e.g. Forest, agriculture & grassland, plantation, sandbar, set­
tlement and waterbody. During 2000–2017, agriculture and grassland were decreasing 23 ha/year. Out of
11,831 ha (in 2000) Agri-grass land 9,326 ha remained intact while remaining 2,246 ha converted to settlement
mixed with homestead plantation class. This study also presents the landscape erosion-accretion due to natural,
quasi-natural and anthropogenic interventions which shows that, along the river flow and at the confluence at
the Nishanbaria Union (local name Khouttar Char & Fakir hat) to lower side of the Tengragiri WS locations are
susceptible to high trend of land erosion whereas accretions are prominent on the reverse sides named Baliatali
Union, Barabagi Union and so on. These results of the study and developed maps will be helpful for the com­
munity people, line departments, national and international policy maker and the researchers’ community for
monitoring coastal geomorphology including erosion and accretion of this landmass.

1. Introduction region and its funnel shape (Rose and Bhaskaran, 2017) expose more
threat to natural disasters for instances cyclone, storm surge, flooding
Coastal environment is a very complex and diverse ecosystem form and so on (Ali, 1996). Shoreline changes also a considerable coastal
the main land (Kirui et al., 2013). Due to climate change and sea level phenomenon in Bangladesh (Salauddin et al., 2018) and Barguna
rise most coastal environment around the world are being affected and Patuakhali Coastal Zone (BPCZ) is situated lower confluence point of
expose threat to its inhabitant (IPCC, 2007). Bangladesh lies in a tropical Ganges-Brahmaputra-Meghna (GBM) basin which is dynamic in

* Corresponding author.
E-mail addresses: [Link]@[Link], mozahid@[Link] (Md.M. Islam), sfr_rahman@[Link] (Md.S. Rahman), alamgirmk@[Link]
(Md.A. Kabir), nazrul_geo@[Link] (Md.N. Islam), ruhulforester@[Link] (R.M. Chowdhury).

[Link]
Received 5 August 2019; Received in revised form 31 October 2019; Accepted 24 November 2019
Available online 27 November 2019
2352-9385/© 2019 Elsevier B.V. All rights reserved.
Md.M. Islam et al. Remote Sensing Applications: Society and Environment 17 (2020) 100277

Fig. 1. Study area.

erosion-accretion pattern. Annual accretion of BPCZ shoreline is 77.087 It is obvious that, earth is changing, and it is an unavoidable natural
m and erosion is 98.26 m (Sarwar and Woodroffe, 2013) whereas it also phenomenon (Van Loon, 2001). Global climate change has largely
massive in some areas. There are about 33.87 million people living in the negative and few positive impacts on population, animal, sea and overall
coastal zone which is considerable i.e. 27% people among total popu­ environment (IPCC, 2007). According to the Gaia Hypothesis, global
lation (BBS & SID, 2011). (see Fig. 1) environment has an ability to recover the natural body of its own or

2
Md.M. Islam et al. Remote Sensing Applications: Society and Environment 17 (2020) 100277

Fig. 2. Overall workflow and methods of this study.

Table 1
Characteristics of the collected Landsat data.
Data Acquired Date Satellite (Sensor) Path/Row Acquisition Time (GMT þ6.00) Resolution (m) Image Quality CC (%)

19-01-2000 Landsat 5 (TM) 137/045 11:19:03 30 � 30 9 0

05-03-2005 Landsat 5 (TM) 137/045 6:36:19 30 � 30 9 2
30-01-2010 Landsat 5 (TM) 137/045 10:32:13 30 � 30 8 0
01-01-2017 Landsat 8 (OLI_TIRS) 137/045 10:46:22 30 � 30 9 4.39

previous status if a change has occurred in a recovery range (Kirchner, 2. Materials and methodology
2002). Despite of those changes, it is required to understand whether it
is in recovery range (Islam and Kitazawa, 2012). Apart from that, Naturally grown mangrove forest Tengragiri wildlife sanctuary
making a place climatically resilience for adapting its existing flora and (locally named ‘Fatrar Bon’) situated under Taltoli sub-district but
fauna, several measures need to take into consideration (Shi et al., associated other mangrove forests grown in part of Patuakhali district so
2018). Before taking the decision to serve any community, it is very far it also considered as area of interest (AOI). Geographically it located
much important to assess the land feature changes over time. In this at 21010 25.7300 N to 210.490 34.75800 N latitude and 900 00 19.54200 E to
regard, LULC change detection study by using satellite imagery is one of 900 90 0.14500 E longitude under Barguna, Patuakhali district in Taltoli
the most useful techniques in recent era (Hussien, 2009). and Kalapara sub district respectively. Taltoli sub district covers an area
Mangroves ecosystem of Bangladesh producing variety of economic, of 258.94 km2 with the population of 88,004 people (BBS & SID, 2011).
ecological services along with green safety barrier from different natural Population density is 541/km2 and literacy rate is 31.76% (BBS & SID,
disasters of the coastal communities (Kathiresan, 2012). Therefore, 2011). Taltoli subdistrict have seven unions and Kalapara subdistrict
coastal area and island including mangrove forest are important have 12 unions. Study area covered total or partially of 5 unions named
candidate for conservation (Kirui et al., 2013). Some studies has taken in Pochakoralia, Chotobogi, Koroibaria, Borobogi, Nishanbaria, Sonakata
broader scale but no specific study has taken focusing mangroves area union (Bangladesh, 2017) of Taltoli subdistrict and four unions named
(Ibrahim et al., 2018) specially in Barguna and Patuakhali. Remote Lata chapali, Kaprabhanga, Nilgonj, Chakamaiya of Kalapara subdistrict
sensing and satellite image analysis based LULC and prediction analysis (under Patuakhali district).
is mostly suitable for intact forest and adjacent areas features identifi­
cation rather than the conventional methods. Using very high resolution
imagery carries information more sophisticatedly and this study has 2.1. Data acquisition and preparation
limitation of using high resolution imagery like worldview-2, RapidEye,
Quickbird and so on but several studies successfully uses Landsat im­ Higher quality images, low cloud cover and lower climatic variable
agery for the relevance work (Iqbal and Khan, 2014; Rawat et al., 2016; satellite imagery collection is the prerequisite for remote sensing-based
Saadat et al., 2011) so far it considered suitable for our study and utmost analysis which is core activity of this study. Freely available data has
tried to conduct sufficient field visits and utilizing other sub-sequent collected form USGS EE website ([Link] Taltoli
logistics for gaining higher accuracy level. To meet the research gaps, sub-district is covering by World Reference System 2 (WRS-2) path 137,
finally 2000, 2005, 2010 and 2017’s Landsat imagery uses for LULC row 045 (Xian et al., 2009). Full study area is covered by this section so
changes with coastal geo-morphological identification of the Taltolli there is no need to make mosaic. Three Landsat 5 TM data were acquired
Sub-district of Barguna and parts of Patuakhali District. in the year 2000, 2005, 2010 and one Landsat 8 data were acquired from
Operational Land Imager (OLI) and Thermal Infrared Sensor (TIRS) in
2017 for LULC change analysis of the studied area and overall workflow
and study methods presented in Fig. 2. In general, accuracy of

3
Md.M. Islam et al. Remote Sensing Applications: Society and Environment 17 (2020) 100277

Table 2
Landscape change rate per annum (CRPA) in the year 2000, 2005, 2010 and 2017.
LULC types 2000 CRPA 00-05 2005 CRPA 05-10 2010 CRPA 10-17 2017 CRPA 00-17

Area (ha) % Area (ha) % Area (ha) % Area (ha) %

Agriculture Grassland 11,848 35% 1.40 11,020 33% 2.02 9,905 30% 3.12 11,452 34% 0.20
Forest 3,306 10% 0.89 3,186 9% 19.87 3,219 10% 35.97 3,759 11% 5.99
Plantation (other) 331 1% 12.08 531 2% 9.30 284 1% 0.77 295 1% 0.64
Sandbar 1,371 4% 10.42 657 2% 21.70 1,370 4% 11.08 611 2% 3.26
Settlement 4,835 15% 7.03 6,535 19% 0.84 6,260 19% 0.17 6,313 19% 1.80
Waterbody 11,873 35% 0.40 11,635 35% 1.53 12,526 37% 2.22 11,134 33% 0.37
Total 33,564 33,564 33,564 33,564

2.3. Field GPS points and ancillary data collection

A pre-classification field visit done in this study where ground

truthing and different time field data acquisition purpose several field
visits were conducted from 20th Jan 2017 to 02nd Feb 2017. During this
period, minimum 30 GPS locations, prevailing features with geotagged
imagery recorded for each class. To collect socio-economic and histor­
ical LULC changes information and magnitudes purpose Focused Group
Discussion (FGD), semi-structural interview, transact work, Training,
also used. CREL’s site office also sufficiently helps to collect data along
with previously conducted forest carbon inventory database. Other
successive secondary database collected from BBS, line ministry, NGO
offices as well.

2.4. Supervised classification and refinement

Fig. 3. Change detection in the year 2000–2005, 2005–2010, 2010–2017 and
overall 2000–2017 interval. Several investigation and testing refer, pixel based supervised clas­
sification using ERDAS Imagine 2014 suitable for this study (Araya and
classification also dependent on different seasons of the year (Xian et al., Hergarten, 2008). Ground controlling sample points collected from field
2019). visit used for creating AOI set and signature files. Maximum likelihood
Due to monsoon climatic condition the study areas optical sensors parametric decision rule has selected and successively mean, Std. devi­
cloud free data is limited. However, dry season (Nov–April) data were ation, lower limit and upper limit also calculate during supervised
collected for different years. Necessary information collected from classification. Post classification digitizing (refinement) used for cor­
Landsat data have been given in Table 1. recting misclassified isolated pixels which required for achieving higher
accuracy of the (Peacock, 2014). After maximum likelihood supervised
2.2. Image pre-processing field visit map preparation classification algorithm, thematic recode tool has used and finally
selected six classes. Clump tool helps to sieve correct isolated pixels and
All composite color bands (except thermal band) have been used minimum 2.00 pixels finds suitable for this study (Intergraph Corpora­
where mosaicking of the imageries not required for selecting our AOI. tion, 2013). Finally, eliminating lower than two pixels and converted
Satellite imagery and GPS locations along with field visited ancillary on-screen digitizing done by using ArcMap 10.5 software.
data were used for this sample selection (Pouncey et al., 1999). To
prepare corrected image, To prepare corrected image, Atmospheric 2.5. Accuracy assessment and map preparation
ATCOR-2 extension developed by the ERDAS Imagine platform used for
atmospheric and haze reduction purpose (Richter and Schla €pfer, 2017). Accuracy assessment considered important for making decision
For radio-metrically image enhancement technique histogram equal­ based on the classified study area map (Brown, 2014). To find the
ization also used and found suitable for this study. The same kind of classified map work accuracy, an updated technology (Brown, 2014)
atmospheric and radiometric correction has done for all four years’ used where minimum 40 points has created per class and verified with
imagery. Unsupervised classification in LULCC analysis purpose found the present extent with the field and historical extent feature of the
necessary in some studies (Were, 2008; Islam et al., 2015)). To do so, google earth pro. Previously collected ground verification points and
K-means clustering method and appropriate true color scheme options associated new field visits also helps in this extent. Using this semi
with 24-time iteration was used which considered best for unsupervised machine learning mechanism found overall accuracy 88.15% and Kappa
classification into several literature (Intergraph Corporation, 2013) to coefficient is 0.867 (in scale 1).
classify the subset into 56 classes in this study. Outcome of the unsu­
pervised classification sorted into 12 major classes and field study map
printed into several A3 and A1 size papers. To identify and incorporate 2.6. Change detection and magnitude of change analysis
recent feature, Google earth-based HR 49 image segments processed and
used. Refer to this activity, Garmin GPSMAP 78s, CASIO QV-R200 Post-classification change detection algorithm provides more accu­
camera was used for identifying appropriate GPS coordinates and geo­ rate result for monitoring and analysis of LULC change detection and
tag image purpose. analysis claimed in several studies (Peacock, 2014). Post-classification
change detection algorithm also provide easy and reliable formula
rather than a complex matrix (Were, 2008). Magnitude of land cover
changes purpose several well-documented literature has adopted which
also used here (Were, 2008)-

4
Md.M. Islam et al. Remote Sensing Applications: Society and Environment 17 (2020) 100277

Fig. 4. Spatio-temporal mapping of the year 2000, 2005, 2010 and 2017.

A2 A1 Δ¼Average annual rate of change (%)

Δ¼ð � 100Þ � ðT2 T1 Þ
A1 A1¼Amount of land cover type in time 1 (T1)
A2 ¼ Amount of land cover type in time 2 (T2)
where.

5
Md.M. Islam et al. Remote Sensing Applications: Society and Environment 17 (2020) 100277

Fig. 5. Natural cause and human interventions for erosion and land cover conversion; first one presents the erosion making trees uprooted, second one present,
fisherman construct temporary hatchet and due to it, side area cleaned and burnt (Source: Field visit 2017–2018).

Fig. 6. Simplified representation of the LULC conversion matrix based on the year 2000–2017. This image presents the inland waterbody is interchanging class with
the sandbars.

3. Result and discussion converted/interchanged into different classes (Table 2). Area change
along with change detection analysis are prerequisite for LULC changes
Study area of interest (AOI) intersects about 33,563 ha in total (Dewan and Yamaguchi, 2009) which presented into Table 2 indicating
where, Agriculture and grassland 12846.1 ha. (38%), Forest 4617.1 ha. units at hector.
(14%), Plantation (other) 1382.4 ha, Sand 494.3 ha, Settlement 2858.3 Table 2 indicates agriculture and grassland area in the year 2000,
ha and Waterbody class covered 11364.8 ha. (34%) respectively 2005, 2010 and 2017 is 11,848 (35%), 11,020 (32.83%), 9,905
extracted from unsupervised classification. LU is a dynamic phenome­ (29.51%) and 11,452 (34.12%) respectively. Fig. 3 indicates the overall
non, which frequently changed by geo-natural process, seasonal agricultural and pasture land has decreased 369 ha from 2000 to 2017,
changes, variation in cropping pattern etc. and so on (Kirui et al., 2013). massive area decreases in 2005–2010 but increased in 2010–2017.
That’s why, total study area will be remained unchanged, but it is Discussing with local community and the stakeholders, they percept two

6
Md.M. Islam et al. Remote Sensing Applications: Society and Environment 17 (2020) 100277

Table 3
Probability cross matrix from 2000 to 2005 shows in hector, the highest intact area are agriculture and waterbody where sandbar changes with time, therefor set­
tlement and built-up area are the second interchanging class.
2000–2005 Agriculture Forest Sandbar Settlement Waterbody Total 2005 (ha)

Agriculture 9001 (0.82) 11 (0) 563 (0.05) 903 (0.08) 542 (0.05) 11,020
Forest 4 (0) 3166 (0.85) 60 (0.02) 271 (0.07) 216 (0.06) 3,716
Sandbar 81 (0.12) 75 (0.11) 301 (0.46) 44 (0.07) 156 (0.24) 657
Settlement 2140 (0.33) 307 (0.05) 220 (0.03) 3443 (0.53) 424 (0.06) 6,535
Waterbody 621 (0.05) 77 (0.01) 228 (0.02) 173 (0.01) 10536 (0.91) 11,635
Total 2000(ha) 11848 (0.35) 3637 (0.11) 1371 (0.04) 4834 (0.14) 11873 (0.35) 33,563

Table 4
Interval of 2005–2010 and area conversion matrix in hector found most of the sandbar converted to sapling area due to planted by forest department which also
showing near reflectance value of the grassland, several areas remains inundated.
2005–2010 Agriculture Forest Sandbar Settlement Waterbody Total 2010(ha)

Agriculture 7682 (0.78) 15 (0) 107 (0.01) 1518 (0.15) 583 (0.06) 9,905
Forest 25 (0.01) 3099 (0.88) 11 (0) 335 (0.1) 32 (0.01) 3,503
Sandbar 711 (0.52) 114 (0.08) 204 (0.15) 177 (0.13) 163 (0.12) 1,370
Settlement 1580 (0.25) 245 (0.04) 50 (0.01) 4056 (0.65) 330 (0.05) 6,260
Waterbody 1021 (0.08) 243 (0.02) 285 (0.02) 449 (0.04) 10526 (0.84) 12,525
Total 2005(ha) 11020 (0.33) 3716 (0.11) 657 (0.02) 6535 (0.19) 11635 (0.35) 33,563

Table 5
Inundated sandbar has recovered, and the sapling converted into seedling area within 2010–2017, but settlement class present the increasing trend.
2010–2017 Agriculture Forest Sandbar Settlement Waterbody Total 2017(ha)

Agriculture 8097 (0.71) 22 (0) 704 (0.06) 1593 (0.14) 1036 (0.09) 11,452
Forest 121 (0.03) 3104 (0.77) 117 (0.03) 509 (0.13) 202 (0.05) 4,054
Sandbar 50 (0.08) 79 (0.13) 151 (0.25) 159 (0.26) 173 (0.28) 611
Settlement 1600 (0.25) 216 (0.03) 216 (0.03) 3936 (0.62) 345 (0.05) 6,313
Waterbody 38 (0) 82 (0.01) 182 (0.02) 62 (0.01) 10770 (0.97) 11,134
Total 2010 (ha) 9905 (0.3) 3503 (0.1) 1370 (0.04) 6260 (0.19) 12526 (0.37) 33,563

Table 6
Overall statistic presents, agricultural area (ha reduced and its converted mostly to Settlement and homestead vegetation, where inland vegetation sometimes mis­
classified with the forest species and waterbody interchanged into subsequent classes.
2000–2017 Agriculture Forest Sandbar Settlement Waterbody Total 2017(ha)

Agriculture 9339 (0.79) 147 (0.01) 57 (0) 2249 (0.19) 56 (0) 11,848
Forest 5 (0) 3057 (0.84) 81 (0.02) 225 (0.06) 270 (0.07) 3,637
Sandbar 555 (0.4) 77 (0.06) 162 (0.12) 289 (0.21) 288 (0.21) 1,371
Settlement 920 (0.19) 511 (0.11) 133 (0.03) 3150 (0.65) 121 (0.02) 4,834
Waterbody 633 (0.05) 262 (0.02) 179 (0.02) 400 (0.03) 10400 (0.88) 11,873
Total 2000 (ha) 11452 (0.34) 4054 (0.12) 611 (0.02) 6313 (0.19) 11134 (0.33) 33,563

severe cyclone named SIDR (2007) and Aila (2009) has occurred in this fruticans) saplings are in severe in danger and not growing as well (see
period. Another study found, about 300 ha landmass inundated in 2010 Fig. 4).
during this imagery capture (30-01-2010), and after that this area Coastal land mass is occupied by large human load due to income
revealed in 2017 satellite imagery. sources and availability of the freely available coastal resources. The
Naturally grown mangrove forest (Tengragiri Wildlife sanctuary) livelihood and the population in the coastal area are constantly
conserves natural food system with some artificial renovations (Tarikul increasing and its always shows increasing sign in our study. It has
Islam and Tribune, 2019). The present food cycle is under severe risk showed pick in 2005 and the figure is 19% which was the same for the
(Staff Reporter, 2019), trails and the other tourism facility need to duration 2000–2017 average. Settlement class has increased 1700 ha
renovate and these are properly documented in different newspapers. In from 2000 to 2005 where it’s damaged by the two super cyclone and
2005, intact forest area reduced about 120 ha which converted to immediate after this image reduced into 275 ha and finally it has in­
sandbar and waterbody (Fig. 6). After declaring this forest as reserve crease and the figure is 6313 ha and it average increased into 87 ha/yr.
forest and plantation by the ministry of forest and other private orga­ Inland waterbody and riverine area like canal and ponds decreasing with
nization forest class providing increasing sign (9.49%, 9.59% and finally time. In the year 2010 waterbody of this area has drastically increased
11.21%). Due to forest plantation and forest regain, overall change line due to inundation cause of super cyclones. Overall it has been decreased
graph (Fig. 3; Green line) providing increasing sign (453 ha) which is 739 ha (44 ha in average) from 2000 to 2017.
11.21% of the total area in 2017.
Plantation forestry considered another LC class due to it is identical 3.1. Land cover conversion analysis
and required for the decision makingFig. 4. Between 2000-2005 about
200 ha sandy area planted where the seedling and saplings area inun­ Study area is intact but interchange into different classes with time.
dated in 2007 and 2010 (Fig. 5). This area is not recovered well which To identify area conversion, study incorporated python based special
also clearly documented during field visit and found, the Goal tree (Nypa programming tools in ArcGIS 10.5 (arcpy). The generated python script

7
Md.M. Islam et al. Remote Sensing Applications: Society and Environment 17 (2020) 100277

Fig. 7. Coastal frontiers accretion-erosion scenario; whereas red color presenting erosion area and light green color presenting area accretion in hector based on the
image of 2000–2017. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article).

shows 50ha less than the real extent so far overlay based conversion tool homestead settlement class. Overall the agricultural land is decreasing
used and found the sandbar is the most interchanging class. day by day and conversely settlement are increasing and the unplanned
changing rate increasing the probable disasters.

3.2. Area conversion and probability matrix

3.3. Coastal frontiers area change analysis
As area changing due to the activity of the human, natural and quasi
natural causes confusion probability matrix with the conversion matrix Coastal land dynamics is a very common natural feature. Lack of
are also analyzed. Transition probability matrix of the interval 2000- siltation tracking procedure, low turbidity from upstream and strong
2005 found sandbar is the most prominent class of changes (Table 3) backwater effect cause for accretion and erosion of this area rapidly. It is
and due to two major natural calamity predicted matrix found settle­ easily identifiable; land cover changes make negative impact for the
ment with homestead class within the interval 2005-2010 is significant communities who are directly dependent on the natural resources
interchanging class (Table 4). Therefore, 2010-2017 found waterbody (Monnereau and Abraham, 2013). Ganges river flow surplus reduced
increasing rate and the probable inundation of the land surfaces pre­ (Rahman and Rahaman, 2018) which may increase the strong backwater
sented in Table 5 where the overall predicted class presented in Table 6 effect on the upper land (Kawser and Abdus Samad, 2016) and fresher
and within this interval significant agricultural land converted to species will migrate (Mansur et al., 2017), so far, the fisheries

8
Md.M. Islam et al. Remote Sensing Applications: Society and Environment 17 (2020) 100277

communities will lose their income sources (Dasgupta et al., 2017). References
Community people claimed, the barrier of Patharghata, Borobogi and
Mithapukur area flood barrier is not protecting them from flood and Ali, a., 1996. Vulnerability of Bangladesh to climate change and sea level rise through
tropical cyclones and storm surges. Water Air Soil Pollut. 92, 171–179. [Link]
associated saline water intrusion (Staff Reporter, 2014). Reference to org/10.1007/BF00175563.
this, the whole communities’ livelihood pattern will be changed (Das­ Araya, Y.H., Hergarten, C., 2008. A comparison of pixel and object-based land cover
gupta et al., 2017) in accordance of the changes of the soil salinity classification: a case study of the Asmara region. Eritrea. WIT Trans. Built Environ.
100, 233–243. [Link]
(Talukder et al., 2012). Bangladesh, G. of, 2017. Taltoli Upazila at a galance. Retrieved from. [Link]
From the (Fig. 7) map, it is clearly identifiable, erosion zone 2 (e 2) [Link]/node/1075311/.
and erosion zone 4 (e 4) are most susceptible to landmark loss, and the BBS, SID, 2011. Population and Housing Census 2011 (Bangladesh).
Brown, K.R., 2014. Accuracy assessment of land-cover maps with high resolution
area loss from 2000 to 2017, 99.97 ha & 245.06 ha respectively. How­ imagery. Winrock Int. 1 (1), 1–13.
ever, most of the area is forestland. Fig. 7 also demonstrates that Ac­ Dasgupta, S., Huq, M., Mustafa, M.G., Sobhan, M.I., Wheeler, D., 2017. The impact of
cretion zone A-1, A-2 & A-4-1 landmark are increasing with time being. aquatic salinization on fish habitats and poor communities in a changing climate:
evidence from Southwest coastal Bangladesh. Ecol. Econ. 139, 128–139. [Link]
Which are respectively 99.90, 57.88 and 51.39 hactor. Though some
org/10.1016/[Link].2017.04.009.
area is increasing frequently but decreasing rate is very high and in Dewan, A.M., Yamaguchi, Y., 2009. Using remote sensing and GIS to detect and monitor
recent past this area losses 300 area. (Source: Forest Beat office, Field land use and land cover change in Dhaka Metropolitan of Bangladesh during 1960 –
survey-2017). Furthermore, community people losing their income 2005, pp. 237–249. [Link]
Hussien, A.O., 2009. Land use and land cover change, drivers and its impact: a
sources and tends to migrate in the divisional town even though in the comparative study from Kuhar Michael and Lenche Dima of Blue Nile and Awash
capital (Mallick and Vogt, 2014), which is exacerbating the pressure to basins of Ethiopia. Retrieved from. [Link]
manage these towns and associate crime profiling is increasing claimed Ibrahim, M., Zaman, M., Mostafizur, A., Shahidullah, S., 2018. Diversity of crops and
land use pattern in Barisal region. Bangladesh Rice J. 21 (2), 57–72. [Link]
by the respective departments (Ishtiaque and Nazem, 2017). To sum up, 10.3329/brj.v21i2.38196.
socio-economic all factors are interlinked so far, it is changing the Intergraph Corporation, 2013. ERDAS Field Guide TM, 2013th ed.
overall income sources, livelihood patterns and sustainable perception IPCC, 2007. Mitigation of climate change: contribution of working group III to the fourth
assessment report of the Intergovernmental Panel on Climate Change. In:
of the community (Nazrul et al., 2018) that’s why we need to address Intergovernmental Panel on Climate Change.
and resolve these problems immediately by taking some engineering Iqbal, M.F., Khan, I.A., 2014. Spatiotemporal land use land cover change analysis and
methods like build super dyke with revised height, groin, repelling erosion risk mapping of Azad Jammu and Kashmir, Pakistan. Egypt J. Remote Sens.
Space Sci. 17 (2), 209–229. [Link]
waterways and so on. Ishtiaque, A., Nazem, N.I., 2017. Household-level disaster-induced losses and
rural–urban migration: experience from world’s one of the most disaster-affected
4. Conclusion countries. Nat. Hazards 86 (1), 315–326. [Link]
2690-5.
Islam, M.N., Kitazawa, D., 2012. Modeling of freshwater wetland management strategies
Study area has economic, enviro-ecological values along with rec­ for building the public awareness at local level in Bangladesh. Mitig. Adapt.
reational benefit and government earning revenues from it so proper Strategies Glob. Change 18 (4), 449–470. [Link]
monitoring and analysis required. Based on this analysis agri-grassland 9369-3.
Islam, M.N., Kitazawa, D., Mannan, M.A., et al., 2015. Modeling Alternative Mitigation
area significantly decreasing about 23 ha/yr, whereas settlement mixed Strategies of Salinity Intrusion on Food Production by Integrating SWOT and TEAM
with homestead plantation continuously increasing in the landmass. Decision Models, Modeling Earth Systems and Environment, vol. 2. Springer
Deep forest, varieties dangerous wild animals and snakes balancing publication.
Kathiresan, K., 2012. Importance of mangrove ecosystem. Int. J. Mar. Sci. [Link]
ecological food chain but it’s being threatening day by day due to org/10.5376/ijms.2012.02.0010.
improper maintenance system. If it continues significant services Kawser, M.A., Abdus Samad, M., 2016. Political history of Farakka Barrage and its effects
including phytoremediation, natural heart and successive carbon on environment in Bangladesh. Bandung: J. Glob. South 3 (1), 1–14. [Link]
10.1186/s40728-015-0027-5.
sequestration in the mangrove forest area will be diminished. The severe Kirchner, J.W., 2002. The Gaia hypothesis: fact, theory, and wishful thinking. Clim.
erosion prone areas named Khouttar char and Fakir hat area needed to Change 52 (4), 391–408. [Link]
take under protection measure. Anyone who can work to protect this Kirui, K.B., Kairo, J.G., Bosire, J., Viergever, K.M., Rudra, S., Huxham, M., Briers, R.A.,
2013. Ocean & Coastal Management Mapping of mangrove forest land cover change
natural mangrove forest and the adjacent landmass this study will be along the Kenya coastline using Landsat imagery. Ocean Coast Manag. 83, 19–24.
helpful, and it is required to protect this threatened naturally grown [Link]
mangrove forest from the national and international level. Mallick, B., Vogt, J., 2014. Population displacement after cyclone and its consequences:
empirical evidence from coastal Bangladesh. Nat. Hazards 73 (2), 191–212. https://
[Link]/10.1007/s11069-013-0803-y.
Declaration of competing interest Mansur, E.F., Smith, B.D., Mowgli, R.M., Diyan, M.A.A., 2017. Two incidents of fishing
gear entanglement of Ganges river dolphins ( Platanista gangetica gangetica ) in
waterways of the. Int. J. Zool. 34 (October 2002), 362–366. [Link]
The authors of this paper and their institutions have no conflicts of
10.1155/2017/8964821.
interest issues. Monnereau, I., Abraham, S., 2013. Loss and damage from coastal erosion in Kosrae, the
The authors are academics affiliated with educational institutions Federated States OF Micronesia. In: The Loss and Damage in Vulnerable Countries
that have no other interests. Initiative.
Nazrul, Islam Md, Al Amin, M., Noman, M., 2018. Management Challenges of sinking of
Oil tanker at Shela Coastal River in Sundarbans mangroves in Bangladesh. Environ.
Acknowledgement Manag. Mar. Ecosyst. 323, 2018 Mar 9.
Peacock, R., 2014. Accuracy Assessment of Supervised and Unsupervised Classification
Using Landsat Imagery of Little Rock. Northwest Missouri State University,
I owe to John a Dhor, Chief of the party, USAID’s CREL project to Arkansas.
give me a chance to keep working in his organization. Sunbeam Rahman Pouncey, R., Swanson, K., Hart, K., 1999. ERDAS Field Guide, fifth ed. Revised and
and Shariful Islam helps me to remote sensing analysis where Mr. Expanded, Imagine.
Rahman, M.M., Rahaman, M.M., 2018. Impacts of Farakka barrage on hydrological flow
Sawpon Chandra, Johir Ahmed’s team helps to collect field data of Ganges river and environment in Bangladesh. Sustain. Water Resour. Manag. 4
collection. Field database and field visit purpose CREL project office (4), 767–780. [Link]
helps me a lot and I am also thankful to the different online and offline Rawat, K.S., Mishra, A.K., Bhattacharyya, R., 2016. Soil erosion risk assessment and
spatial mapping using LANDSAT-7 ETMþ, RUSLE, and GIS—a case study. Arab. J.
portals for helping me in technical part of this research work. Geosci. 9 (4) [Link]
Richter, R., Schl€apfer, D., 2017. Atmospheric/topographic correction for satellite
Appendix A. Supplementary data imagery (ATCOR - 2/3 user guide). In: ATCOR-2/3 User Guide, pp. 1–71, Version
9.1.2.
Rose, L., Bhaskaran, P.K., 2017. Tidal propagation and its non-linear characteristics in
Supplementary data to this article can be found online at [Link] the head Bay of Bengal. Estuar. Coast Shelf Sci. 188, 181–198. [Link]
org/10.1016/[Link].2019.100277. 10.1016/[Link].2017.02.024.

9
Md.M. Islam et al. Remote Sensing Applications: Society and Environment 17 (2020) 100277

Saadat, H., Adamowski, J., Bonnell, R., Sharifi, F., Namdar, M., Ale-Ebrahim, S., 2011. Staff Reporter, T., 2019. Tengragiri Eco-Park Wildlife Has Dearth of Food. Prothom Alo,
Land use and land cover classification over a large area in Iran based on single date pp. 1–2.
analysis of satellite imagery. ISPRS J. Photogrammetry Remote Sens. 66 (5), Talukder, N.I., Rahman, M., Rafiq, G., Ahmed, U., Gharami, G., 2012, February 2. Sorrow
608–619. [Link] of 30 Barguna Villages. The Daily Star, pp. 0–3.
Salauddin, M., Hossain, K.T., Tanim, I.A., Kabir, M.A., Saddam, M.H., 2018. Modeling Tarikul Islam, M., Tribune, D., 2019. Tengragiri Eco-Park Needs Renovation. Dhaka
spatio-temporal shoreline shifting of a coastal island in Bangladesh using geospatial Tribune, pp. 1–2.
techniques and DSAS extension. Annals of Valahia University of Targoviste. Geogr. Van Loon, A.J., 2001. Changing the face of the earth. Earth Sci. Rev. 52 (4), 371–379.
Ser. 18 (1), 1–13. [Link] [Link]
Sarwar, M.G.M., Woodroffe, C.D., 2013. Rates of shoreline change along the coast of Were, K., 2008. Monitoring Spatio-Temporal Dynamics of Land Cover Changes in Lake
Bangladesh. J. Coast. Conserv. 17 (3), 515–526. [Link] Naivasha Drainage Basin. International Institute for Geo-Information Science and,
013-0251-6. Kenya. Retrieved from. [Link]
Shi, H., Rigge, M., Homer, C.G., Xian, G., Meyer, D.K., Bunde, B., 2018. Historical cover Xian, G., Homer, C., Fry, J., 2009. Updating the 2001 National Land Cover Database land
trends in a Sagebrush Steppe ecosystem from 1985 to 2013: links with climate, cover classification to 2006 by using Landsat imagery change detection methods.
disturbance, and management. Ecosystems 21 (5), 913–929. [Link] Remote Sens. Environ. 113 (6), 1133–1147. [Link]
10.1007/s10021-017-0191-3. rse.2009.02.004.
Staff Reporter, T., 2014. Flood control dams collapse. The New Nation. July 16, pp. 2–3. Xian, G., Shi, H., Dewitz, J., Wu, Z., 2019. Performances of WorldView 3, Sentinel 2, and
Retrieved from. [Link] Landsat 8 data in mapping impervious surface. Remote Sens. Appl.: Soc. Environ. 15
-[Link]. (June), 100246. [Link]

10