2 (2018), 2, 87-95

Remote Sensing of Land

2582-3019

Identifying Most Influential Land Use Parameters Contributing Reduction of Surface Water Bodies in Rajshahi City Bangladesh: A Remote Sensing Approach

Abdulla - Al Kafy 1 , Muhaiminul Islam 2 , ABDUR ROUF KHAN 2 , Lamia Ferdous 2 , MD. MAMUN HOSSAIN 2

1.ICLEI South Asia, Rajshahi City Corporation, Rajshahi-6203.

2.Department of Urban and Regional Planning, Rajshahi University of Engineering and Technology (RUET), Rajshahi-6204, Bangladesh.
Mr.Abdulla - Al Kafy*

*.ICLEI South Asia, Rajshahi City Corporation, Rajshahi-6203.

Dr.Suresh Kumar 1

1.Agriculture, Forestry & Ecology Group, Indian Institute of Remote Sensing, Dehra Dun -248001, Uttarakhand, India.

https://doi.org/10.21523/gcj1.18020202
14-04-2019
18-02-2019
26-03-2019
05-04-2019

Graphical Abstract

Highlights

  1. Surface water bodies are reducing at rapid rate in Rajshahi City of Bangladesh.
  2. Remote sensing and GIS technologies were used to estimate the water body reduction.
  3. Around 14% of water bodies were filled up in 25 years due to rapid urbanization in Rajshahi City.
  4. Findings are useful to the policy and decision maker for sustainable surface water management.

Abstract

Surface water bodies are one of the irreplaceable natural resources for human survival, and it extensively reduces with increasing the world population. This study modeled the spatiotemporal changes of land use / land cover (LULC) and identified the most influential LULC parameters, which contributes in the reduction of surface water bodies using the Landsat 4 and 5 TM and Landsat 8 OLI images (1992-2017). Rajshahi City Corporation is situated in the Northern piece of Bangladesh. A maximum likelihood supervised images classification algorithm was used for detection of changes in LULC. Matrix union technique was used for identifying the prominent LULC parameters. About 14% of water bodies were filled up in twenty-five year (1992-2017) due to rapid urbanization in Rajshahi City Corporation area. This study can provide an essential move towards necessary actions for preservation of surface water bodies to maintain the ecological balance and environmental sustainability.

Keywords

Rajshahi , Preservation , Urbanization , Change Detection , LULC , Remote Sensing , Water bodies

1 . INTRODUCTION

Metropolises are engines of development. The dynamic development of urban areas occurs because of the unavoidable changes allocated by many land use /land cover (LULC) parameters (Aboelnour and Engel, 2018; Rahman et al., 2018). LULC change examinations one of the huge apparatuses to survey worldwide change in different spatial-temporal scales (Lambin, 1997; Lambin and Ehrlich, 1997). It reflects the complex interactions of dimensions of human activities and its impact on environment (Xiao et al., 2006). LULC change configurations are depend on natural and socioeconomic expansions accelerated by space and time (Aboelnour and Engel, 2018; Allam et al., 2019).

One of the major parameters for LULC change is rapid urbanization, especially in a developing country. The expansion in population prompts the weights on urban region, farming area, and water assets which consequently subsidizes to LULC changes (Msofe et al., 2019; Sharma, 2019).

Past few decades have witnessed many changes at global scale. The global water resources consumption will be increased by 1.33 times in 2025 when compared to 2070 km3/year in 1995 (Yan et al., 2019). In the past decades, the temperatures have been constantly mounting across the global temperature, and rainfall and precipitation patterns have become unpredictable. The available water resources have become gradually limited; with the economic developments and rising population the consumption and usage of the water resources is continuously increasing (Li and Qian, 2018; Liu et al., 2019). The sustainable management of the available water resources is thus of prime importance. Water is an essential natural resource with limited availability. Water bodies play an essential role in the ecological environment and LULC change significantly impact on surface water bodies in rapid urbanized area (Li and Qian, 2018; Liu et al., 2019; Yan et al., 2019). This sort of changes in water bodies extraordinarily influences the neighborhood and local conditions, which would in the long run influence on the worldwide condition. Substantial reduction of surface water bodies mostly because of human-induced LULC changes impacting the worldwide carbon cycle and adding environmental CO2 (Alves and Skole, 1996; Dannenberg et al., 2018). It is therefore fundamental to recognize most persuasive LULC parameter contribute to reduction of surface water bodies with the goal that its impact on the earthbound biological community can be observed, and main LULC parameters can be defined (Goudie, 2018; Muttitanon and Tripathi, 2005).

The massive demographic development and rapid urbanization causes serious problem to the natural resources especially in developing countries like Bangladesh (Kafy, 2018; Rahman et al., 2018). Rajshahi is one of the largest metropolitan city of Bangladesh currently facing uncontrollable urban growth. The developing cities are also facing urban sprawl which causes an adverse effect on surface water bodies and environmental sustainability (Kafy, 2018). Different land cover change results in declination of surface water bodies by increasing the impervious surface as well as the built-up area (Faridatul and Jahan, 2014). In the last decades, Remote Sensing (RS) and GIS techniques have been used extensively for classifying land cover changes. This classification technique is very helpful to detect the water bodies demolishment resultant from rapid increase in different LULC (Ahmed and Dewan, 2017; Kafy, 2018; Kafy and Ferdous, 2018; Kafy et al. 2018; Mahmud et al., 2011).

Different Geographic Information System (GIS) and Remote Sensing (RS) applications to water resources have achieved throughout the years, and the applications effectively light up the probability of GIS and RS innovation (Rahaman et al., 2018). Ahmed et al. (2013) used Landsat TM and ETM+ to monitor urban growth and temperature changes in the capital city of Dhaka (Ahmed et al., 2013). Mahmud et al. (2011) have identified the spatiotemporal changes of wetland in Dhaka City using GIS and RS (Mahmud et al. 2011). Dewan and Corner (2013) write a book on geospatial perspectives on urbanization, environment and health based on Dhaka megacity (Dewan and Corner, 2013). The information acquired from GIS and RS application provides an updated inventory about the previous and current LULC resources that exist and used by the human (Cyr-Gagnon and Rodriguez, 2019; Rahaman et al., 2018; Singh, 2018; Yan et al., 2019).

This study aims at an integration of RS datasets and GIS techniques to detect the decadal changes in water bodies of Rajshahi City Corporation (RCC) and describes the importance of surface water bodies to establish RCC more sustainable in the near future. The study illustrates the extraction of the persistence of the water bodies over a time using Remote Sensing (RS) images. The application of Geographic Information System (GIS) is helpful to understand the decreasing pattern of water bodies which supports in decision making and achieve sustainable communities in Bangladesh.

2 . STUDY AREA

Rajshahi is one of the biggest metropolitan city of Bangladesh which lies between 24°07′ to 24°43′N and 88°17′ to 88°58′E (BBS, 2011). It was one of the first Municipalities in Bangladesh, established in 1876 and declared as a City Corporation (48.02 km2) in 1987 (BBS, 2013). The region comprises of Barind tract, Diara and Char lands (Figure 1). Rajshahi City situated on the bank of waterway Padma (Clemett et al., 2006; Faridatul and Jahan, 2014). The atmosphere of the region is tropical and it encounters three characterized seasons: (1) sweltering climate from February to June, (2) stormy season from July to October and (3) winter season from November to January. The summer is a lot rainier than the winter in Rajshahi (Ferdous and Baten, 2011). The normal yearly temperature in Rajshahi is 25.8°C with 1419mm of precipitation yearly (Aboelnour and Engel, 2018; Ayub and Miah, 2011; Mahmud et al., 2011). Geological stability encompasses hard rocks having unfavorable hydraulic pressure characterized by poor to moderate groundwater potentials (Haque et al., 2012; Rasel, et al., 2015). The number of water bodies decreased year after years due to urbanization in RCC area (Kafy, 2018; Kafy et al., 2018). 373 ponds were observed in the year 2011 (Kafy et al., 2018; BBS, 2013). Most of the wetland areas were converted into the built-up area which negatively affected the environmental suitability (Kafy et al., 2018).

 

Figure 1. Location of Rajshahi City Corporation (RCC) area: (a) Bangladesh and Rajshahi District and b) RCC ward boundary and LULC

 

3 . DATA AND METHODOLOGY

This study deals with various multi-temporal Landsat 4 and 5 TM and Landsat 8 OLI satellite images (30m resolution) gained from the USGS website (https://earthexplorer.usgs.gov) for the years 1992, 1997, 2002, 2007, 2012 and 2017. The images were captured in cloud free late autumn (September, October and November) (Kafy, 2018; Kafy and Ferdous, 2018) and used to prepare LULC maps. The images were processed in ERDAS Imagine 15 software and final outputs mapped in ArcMap 10.6. The cloud coverage of these images was less than 10% and useful to avoid the atmospheric correction that could affect the accuracy of the classification (Allam et al., 2019; Foody, 2002). All images were geometrically corrected to WGS84 of UTM zone 46N. All these images were captured with maximum one month interval for six different years to avoid the seasonal variation (Kafy et al., 2017).

Band 1 of Landsat 4 and 5 TM satellite and Band 1, 2 and 8 of Landsat 8 OLI images makes noise in the image and create difficulties in taking signature for LULC classification (Estoque and Murayama, 2015; Fisher et al., 2016; Jia et al., 2014; Zhu et al., 2016). Therefore, these bands have not been taken into image processing for change detection of water bodies in the study area (Kafy et al., 2017).

3.1 Image Classification

Four broad categories of LULC (water body, built-up area, vegetation and agricultural land and bare soil) are selected to map the study area. LULC maps of the RCC were prepared for 1992, 1997, 2002, 2007, 2012 and 2017 using  Maximum Likelihood Supervised Classification(MLSC) technique in ERDAS Imagine 15 software. Each classified map evaluated for accuracy assessment using available field data and Google Earth images over randomly selected 350 points (Table 1) (Kafy, 2018; Kafy et al., 2017).

Classification accuracy

User Accuracy (%)

Producer Accuracy (%)

 

Year

Water body

Built-up area

Vegetation

Bare soil

Water body

Built-up area

Vegetation

Bare soil

Overall Accuracy

(%)

Overall

Kappa Statistics

1992

100

75

90

95

100

100

100

67

90

0.8887

1997

100

85

85

80

100

100

71

100

90

0.9001

2002

100

80

80

80

100

100

83

100

93

0.9333

2007

100

75

100

85

100

100

100

67

88

0.8767

2012

100

100

90

80

100

100

71

100

88

0.8800

2017

100

85

100

80

100

100

83

100

91

0.9133
 

3.2 Change Dection using Matrix Union

A change detection technique should be appropriate for the study to identify the water body fill up locations associated with major LULC changes. Application of matrix union tool estabilshed in ERDAS Imagine 15 software. This tool is helpful to identify the transformation of number of pixels from one class to another class. Conversion of water body to built-up area, vegetation and agricultural land and bare land classes were estimated after post-classification change in LULC between 1992-1997, 1997-2002, 2002-2007, 2007-2012, 2012-2017 and 1992-2017. These changes were shown in six maps. The locations of water body fill-up in respect of increase of built-up area, vegetation and agricultural land and bare land were detected and mapped in Arc GIS 10.6.

4 . RESULTS AND DISCUSSIONS

4.1 LULC Change Analysis

Changes in LULC (1992, 1997, 2002, 2007, 2012 and 2017) can be easily determined using classified images. Water bodies are reducing gradually and replaced by built-up area, vegetation and agricultural land and bare land during the urbanization process of Rajshahi city to fulfill the needs of rapid growth of the city and population (Figure 2). Total water body in Rajshahi was 6.12 km2 in 1992 and urban area was 8.8542 km2. A little change over the water body was noticed in 2006 which reduced to 4.4991km2. A significant change observed in a last decade. Human settlement encroached upon water body and reduced it to 2.0151 km2. Urban area increased to 16.6716 km2, which is double than the year of 1996.

 

Figure 2. Land use land cover map of Rajshahi City Corporation in 1992 (a), 1997(b), 2002(c), 2007(d), 2012(e) and 2017(f)

 

The accuracy of classified maps were estimated to verify reliability and acceptance (Allam et al., 2019; Owojori and Xie, 2005). It was estimated based on field check data and Google Earth images. Information about 350 ground truth data points were collected in May 2017 using the stratified random sampling techniques. The reference data and classified images were compared to estimate user’s and producer’s accuracy (Table 1) (Bakr et al., 2010). Non-parametric Kappa statistics also calculated using ERDAS imagine 15 software. Accuracy was estimated about 85% with no individual class less than 70% for LULC classes mapped using Landsat data, (Foody, 2002; Story and Congalton, 1986). The estimated accuracy of the classified images for six years ranges from 87 to 91 % representing very good results.

4.2 Conversion of Water Bodies into Different Influential LULC

4.2.1 Five years interval

LULC changes from 1992 to 2017 are significantly different in dissimilar areas of water body, built up area, vegetation and agricultural land and bare lands (Figure 3). The categories: vegetation and agricultural land and water body were reduces from 1992 to 2017 and the built-up area increased due to upsurge of urbanization. The bare lands are also suppressing show larger amount.

 

Figure 3. Change of area in different LULC

 

The water bodies are filling up due to conversion of land of RCC area for estimated with five years interval (1992-1997, 1997-2002, 2002-2007, 2007-2012 and 2012-2017) (Figure 4). The filled up areas are converted into built-up area, vegetation and agriculture and bare lands.

 

Figure 4. Water body fill up for influential LULC parameters (five years interval): a) 1992-1997, b) 1997-2002, c) 2002-2007, d) 2007-2012 and e) 2012-2017

 

Table 2. Loss of water bodies

 

Change (%) in different years interval

 

Water bodies converted into

Ten years

Twenty years

Twenty five years

1992-2002

2002-2012

2007-2017

1992-2012

1997-2017

1992-2017

Built-up area

4.72

1.82

4.30

4.13

4.89

13.84

Vegetation and Agricultural Land

1.77

0.35

0.81

0.56

2.03

3.82

Bare land

3.30

1.36

3.23

2.76

3.77

10.46

Total

9.79

3.53

8.34

7.45

10.69

28.12
 

 

In 1992-1997, water bodies are converted mostly into built-up areas (3.4%) (Figure 4a), in 1997-2002, converted into built-up area and vegetation and agriculture (1.4%) (Figure 4b). Water bodies are converted probably into built-up areas estimated about 1% in 2002-2007 (Figure 4c), (0.8%) in 2007-2012 (Figure 4d) and 3.5% in 2012-2017 (Figure 4f). Most of the conversions of the water bodies were observed into built-up area in last few years (Figure 5); 1.73 sq. km in 2012-17 and 1.64 sq. km in 1992-1997 (Figure 5).

 

Figure 4. Water body fill up for influential LULC parameters (five years interval): a) 1992-1997, b) 1997-2002, c) 2002-2007, d) 2007-2012 and e) 2012-2017

 

Figure 5. Conversion of water bodies into influential LULC parameters (five years interval)

 

 

4.2.2 Ten years interval

The increase in built-up area was observed about 2.32 sq. km in 1992-2002 and 2.10 sq. km in 2007-2017 (Figure 6). The bare lands are also increased. Therefore, both manmade and natural causes are responsible for filling up the ponds.

 

Figure 6. Conversion of water bodies into influential LULC parameters (ten years interval)

 

Figure 7. Conversion of water bodies into influential LULC parameters (twenty years interval)

 

4.2.3Twenty years interval

The visible changes are noticed in conversion of built-up area, vegetation and agricultural land and bare land from water body in twenty years interval. Despite of the increase of built-up area and bare land (described as the above 5 years and 10 years interval), the increase level of vegetation and agricultural land is also noticed (0.28 sq. km to 0.99 sq. km) (Table 2). Perhaps it can be estimated that the water bodies are filled up by the farmers to grow crops.

4.2.4 Total loss of water bodies

Water body losses were estimated using RS and GIS techniques for 25 years. Estimated converted area of water bodies to built-up areas is 13.84% (Figure 8) areas due to rapid urbanization without superiority on conservation rules. The conversion of water body into bare land is 10.46% (3.5 km2) of the total area (Table 3) due to natural changes. About 3.82% water bodies are converted to vegetation and agricultural land.

 

Figure 8. Conversion of water bodies into influential LULC parameter in 25 years

 

Table 3. Loss of water bodies in 25 years

Water bodies converted into

Area losses

km2

%

Built-up area

6.8

13.8

Vegetation and agricultural land

1.9

3.8

Bare land

5.1

10.5

Total

13.8

28.1
 

 

 

5 . CONCLUSION

The study shows that rapid urbanization is the main cause of water body demolition in the study area. Water bodies are critical component to enhance the quality of the physical and social environment of the city. It carries significant aesthetic and environmental value by providing fresh air in its surrounding areas. The water bodies are often seen as a burden and constraint in the process of growth and development of cities and towns in recent years and getting less importance and transforming into other parameters of LULC. This study illustrates the spatiotemporal changes of land cover and identified the most influential LULC parameter which contributes to the reduction of the surface water body. About 13.84% of water bodies were converted into built-up area during the last twenty five years due to urbanization. Biodiversity damage, severe environmental humiliation, water logging, metropolitan flooding and loss of valuable water resources are the serious effects of water body losses. In order to improve the ecological balance of the RCC area, surface water bodies should be conserved and scenic views around the water bodies need to be established to create more attraction for urban dwellers.

This study provides reliable LULC change analysis useful and necessary in the subsequent revision of biodiversity conservation and policy implementation. Government and policymakers need to take essential measures to protect the water bodies from being converted into different LULC parameters. Graduation and post-graduation academic curriculum should also be reinforced by including topics such as water resource management and planning, needs of recent environmental challenges faced by water resources, etc. Finally, enforcement of water related regulations and strengthening environmental awareness will be helpful to reduce the loss of water bodies. This study will also be useful for policy maker to bring effective initiative by providing adequate insight regarding the pattern and reason for water body conversion. Both the local government and people need to join hands to solve the issues pertaining to water bodies.

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgements

Authors are thankful to the United States Geological Survey (USGS) for Landsat images.

Abbreviations

GIS: Geographic Information System; LULC: Land Use / Land Cover; MLSC: Maximum Likelihood Supervised Classification; RCC: Rajshahi City Corporation; RS: Remote Sensing; TCC: True Color Composite.

References

6.

Ayub, R. and Miah, M. M., 2011. Effects of change in temperature on reference crop evapotranspiration (ETo) in the northwest region of Bangladesh. Paper presented at the 4th Annual Paper Meet and 1st Civil Engineering Congress, December.

8.

BBS [Bangladesh Bureau of Statistics], 2011. Statistical Yearbook of Bangladesh. Statistics Division, Ministry of Planning, Dhaka, Government of the People’s Republic of Bangladesh.

9.

BBS [Bangladesh Bureau of Statistics], 2013. District Statistics 2013, Rajshahi. Bangladesh Bureau of Statistics.

10.

Clemett, A., Amin, M. M., Ara, S. and Akan, M. M. R., 2006. Background information for Rajshahi City, Bangladesh. WASPA Asia Project Report, 2, 1-29.

13.

Dewan, A. and Corner, R., 2013. Dhaka megacity: Geospatial perspectives on urbanisation, environment and health: Springer Science and Business Media.

15.

Faridatul, I. and Jahan, S., 2014. People’s perception regarding the development of community facilities: A case study of Rajshahi City Corporation. J. Bangladesh Inst. Plan. 7, 1-16.

19.

Goudie, A. S., 2018. Human impact on the natural environment. John Wiley and Sons.

22.

Kafy, A. and Ferdous, L., 2018. Pond Filling Locations Identification Using Landsat-8 Images In Comilla District, Bangladesh. Paper presented at the 1st National Conference on Water Resources Engineering (NCWRE 2018) CUET, Chittagong, Bangladesh.

24.

Kafy, A., Ferdous, L., Faisal, A., Khan, H. and Sheel, P., 2018. Exploring The Association of Surface Water Body Change and Rapid Urbanization in Rajshahi City Corporation (RCC) Area Using RS and GIS. Paper presented at the 1st National Conference on Water Resources Engineering (NCWRE 2018).

25.

Kafy, A., Rahman, M. S., and Ferdous, L., 2017. Exploring the association of land cover change and landslides in the Chittagong hill tracts (CHT): A remote sensing perspective.

33.

Owojori, A., and Xie, H., 2005. Landsat image-based LULC changes of San Antonio, Texas using advanced atmospheric correction and object-oriented image analysis approaches. Paper presented at the 5th international symposium on remote sensing of urban areas, Tempe, AZ.

36.

Rasel, H., Hasan, M., Ahmed, B. and Miah, M., 2015. Assessment of Ground Water fluctuation and Recharge due to rainfall in Barind Area under Greater Rajshahi District (North Western Part of Bangladesh). International Journal of Civil and Environmental Engineering IJCEE-IJENS, 13(02), 15-25.

37.

Sharma, K., 2019. Urbanization induced land use-land cover changes in the Manipur valley and surrounding hills: A landscape metrics approach. Environmental Change in the Himalayan Region, 137-155, Springer.