Introduction

The developing countries in the world have been facing rapid urban population growth and urbanization which is the foremost environmental challenge for governments and planners. This predictable process is expanding within the city, and then crossed beyond the city limits, and goes into the hinterland that should need to adopt effective sustainable planning (Mosamman et al. 2017). Land-use change analyses provide evidence of land and give the clear prospects of spatial patterns, trends and rate of impact on the environment which would help to make ecological planning at the regional level as well as good governance for each spatial unit (Ramachandra et al. 2012). United Nations Reported (UN Report) has reported that 54% of the world population lives in urban areas and it is predicted to cross 6.3 billion in 2050. The developing world cities would be reached with approximately 90% of the urban population soon (UN Report 2015). Especially in African and southeastern Asian countries will be experienced more urban population growth. Moreover, urbanization is a process of conversion of land from rural to urban through migration, socio-economic development, opportunities and the number of facilities increases (Taubenbock et al. 2009; Bhagat and Mohanty 2009). Urban sprawl is defined as a dispersion of spatial expansion of the city or town towards its peripheral and sub-urban areas (Ewing 1997; Galster et al. 2001; Tewolde and Cabral 2011). The causes of urban sprawl, i.e., population growth, economic development, an extension of roads and highway networks, open up huge space from the core city, ribbon development in urban structure, etc.(Ewing 1997; Galster et al. 2001; Sudhira et al. 2003). Three ways of urban sprawl development are (1) ribbon sprawl, (2) leapfrog sprawl and (3) low density (radial) sprawl (Falah et al. 2019). As consequences of globalization, privatization and liberalization, the Indian megacities (like Kolkata, Mumbai, Delhi, Bangalore, etc.) have rapidly expanded in the form of sprawl to accommodate the migrants from rural areas and encroaches adjacent smaller towns and villages (Shaw and Satish 2007). Urban sprawling could impact ecological habitat with increases in impervious areas (Liu and Phinn 2001). Urbanization and land use and land cover change (LULCC)are positively related to the negative impact on the environment (Mosamman et al. 2017). Expanded urbanized areas, environmentally impact on surrounding valuable natural lands by encroaches of agricultural fields, forest lands or wetlands (Pathan and Jothimani 1989; Xu et al. 2000; Kumar et al. 2007), the loss of productive land and biodiversity (McKinney 2002; Atu et al. 2013), degradation of air quality, groundwater quality (Al-Kharabsheh and Ta'any 2003; Poyil and Misra 2015), changes of climatic scenario, urban flooding, increases runoff, etc. (Miller and Hutchins 2017; Patra et al. 2018). Urban sprawl has impacts on socioeconomic dimensions in a way of decline a sense of community and cultural value, reducing the health of people, loss of public spaces, traffic issues (Ewing et al. 1997; Hathout 2002; Bento et al. 2003; Carruthers and Ulfarsson 2003; Nechyba and Walsh 2004; Resnik 2010; Jaeger et al. 2010; Pereira et al. 2014). Monitoring of spatio-temporal land-use changes can help the administration and planners to understand city growth and can make a better plan to build a city in sustainable manners (Ramachandra et al. 2014a, b).

Different new approaches or techniques have been introduced in the last few decades to analyze and monitoring of sustainable urban structure, the form of the city, urban sprawl studies over the spatial context (Mosamman et al. 2017). Based on the indicators some researchers have done sprawl analysis (Burchel et al. 1998; Malpezzi1999; Galster et al. 2001; Ewing et al. 2002; Hasse and Lathrop 2003).On the other hand, remote sensing (RS)and GIS techniques with statistical analysis have been used to measure sprawl  at the regional level by different researchers and scholars (Sarvestani et al. 2011; Hafez 2011; Belal and Moghanm 2011; Rawat and Kumar 2015; Poyil and Misra 2015; Hassan et al. 2016; Subasinghe et al. 2016; Kim 2016; Hassan 2017; Mas et al. 2017; Mosamman et al. 2017; Sahana et al. 2018; Shukla and Jain 2019; Girma et al. 2019).

RS and GIS techniques are a tool and have the flexibility for measuring urban sprawl and LULCC through modeling using satellite imageries from micro to global scale (Pathan and Jothimani 1989; Pathan et al. 1993; Donnay et al. 2001; Zha et al. 2003; Lu and Weng 2005; Deng et al. 2008; Parker et al. 2008). In this technique, multi-resolution data provides a clear picture of the landscape (Lillesand and Keifer1987). This is useful for mapping of urban expansion and prediction of LULCC over the different periods (Hathout 2002; Mondal et al. 2016; Mosamman et al. 2017; Liping et al. 2018; Falah et al. 2019). Several new models and approaches have also been introduced for sprawling characteristics in the urban landscape, such as Cellular Automata (CA) modeling algorithm (Li et al. 2014a, b; Mondal et al. 2016), logistic regression model (Hu and Lo 2007; Karolien et al. 2012; Sarkar and Chouhan 2020), Markov Chain (Alsharif and Pradhan 2014; Bose and Roy Chowdhury 2020), CA-based SLEUTH (Clarke et al.1996; Clarke and Hoppen 1997; Chaudhuri and Clarke 2019), geographical weighted regression (Mondal et al. 2015a, b), Artificial neural network (ANN) algorithm (Pijanowski et al. 2005; Mohammady and Delavar 2016), Fuzzy logic approach (Liu and Phinn 2001) and so on. Shannon’s entropy model has been introduced in recent days for urban growth analysis(Sudhira et al. 2003, 2004; Batty et al. 2014; Rahaman et al. 2018; Yulianto et al. 2019; Ismael 2020). The urban structure has been studied using landscape metrics and modeling in geospatial sciences. Landscape metrics modeling was used by several scholars over different cities worldwide (Ramachandra et al. 2012; Ramachandra et al. 2014a, b; Ramachandra et al. 2014a, b; Akin and Erdogan 2020). Urban Sprawl measurement is a tabular formation and quantitative approach to visualize a map of space phenomenon between few and far (Das Chatterjee et al.2016). Through this analysis, it can also be quantified and computes the pattern of urban dynamics and different land features in the city (McGarigal and Marks11995). Landscape metrics define patch area, percentage of landscape (PLAND), largest patch index (LPI), fractal dimension index (FRAC), shape index (SHAPE) and ecological sustainability as well (Dewan et al. 2012).

In India, the cities are sprawling, very rapidly. According to the census 2011 report, the urban population is 377.1 million. It was added 9.1 crore during the last 10 years. It was the first time that the absolute number of urban populations has increased than the rural population. The rural–urban population distributions were 68.84% and 31.16%. The level of Urbanism has also been increased by 31.16% in the Census of India 2011and the rural population decreased from 72.19 to 68.84%. The growth of the urban population has led to the growth of urban land due to mainly migration (Ramachandra and Uttman 2008).

Nationwide, several studies have been done on LULCCand urban growth monitoring in several cities of India. Sudhira et al. (2003) used Shannon’s entropy index to measure sprawl in Karnataka areas of Mangalore and Udupi coastal cities. Ramachandra et al. (2012) have explained the urban dynamics in Bangalore city using landscape metrics to get depth known about sprawl. The urban gradient and sprawl were measured using Shannon’s entropy, which confirms the city growth with dispersed manners. PCA was used for metrics prioritization in this study. Bhatta and Giri (2012) simulated the urban growth over 30 years and predicted future growth up to 2050 of the Kolkata Metropolitan Authority (KMA). Rawat and Kumar (2015) have monitored land use and cover changes in the Hawalbagh block level of Uttarakhand using supervised classification methods. Poyil and Misra (2015) focused on quantitative studies of urban agglomeration in the Malegaon district. The morphological growth of the study area has analyzed using satellite imageries. Mondal et al. (2015a, b) in their article examined the relative changes of each land uses through the calculation of gross gain, gross losses, persistence and net changes in the Kamrup Metropolitan district of Assam. Das Chatterjee et al. (2016) have been carried out the research on Bhubaneswar city using spatial modeling in the measurement and monitoring of land-use changes. The landscape metrics, entropy methods and principal component analysis were used for computing the urban sprawl. Mondal et al. (2016) predicted the future urban sprawl of Kolkata metropolitan areas through using Markov chain simulation. The additional factors (locational force, land use force, market forces, residential force and a group of resistance force) were selected for an explored upcoming urban scenario. Sahana et al. (2018) and Rahaman et al. (2018) analyzed spatial urban sprawl matrix in Kolkata urban agglomeration over the periods of 1990–2015. Shukla and Jain (2019) in their research paper analyzed the process of urbanization of urban, sub-urban, sub-rural and rural levels over the Lucknow city. The pixel-based classification of urbanness or urban densities were categorized into seven classes in the study area. Four types of sprawl are identified and mapped, i.e., infill, extension, leapfrog and ribbon expansion.

Apart from this, the Barrackpore is a sub-divisional area of North 24 Parganas district, West Bengal which contains 16 municipal towns or ULBs (Urban Local Bodies). Industrial locations along the Hooghly river had caused to grow densely population among the municipalities. These urban bodies are under the KMDA (Kolkata Metropolitan Development Authority). A more than a few studies have done on urban growth at KMDA but those were considered the Kolkata city as a core study of urban sprawl in a spatial manner. The aim of the study is to the monitoring of urban sprawl of these towns (municipalities) can support to take suitable planning and minimizing the environmental problem at the micro-level, such as microclimate changes, ecosystem fragmentation, water bodies management, proper urban structure, urban green space management, sustainable livelihood. The present study has first analyzed and detected the pattern of land-use changes in temporal order (1972, 2001, 2016) using remote sensing data in every town. The landscape metrics were used to capture the dynamism of urban land surface changes and growth consideration and lastly, to confirm the sprawl, Shannon’s entropy model was computed to understand with details the local pattern of growth in the UBLs of the Barrackpore sub-division. The micro-scale study is necessary to develop socioeconomic conditions and need to build appropriate management for the small spatial unit.

Study area

The Barrackpore is one of the old sub-divisional parts of North 24 Parganas, West Bengal. It has historical importance due to colonization and industrial influences in the eastern part of India. Physical settings of the sub-division reveal that it is part of the Bengal Basin. The bank of the Hooghly River lies on the right side of the study area. The administration set up of the sub-division at present is 16 municipalities or ULBs (Urban Local Bodies), 2 CD block, 24 census town 53 Villages, 14 g Panchayat and 13 police stations (Census of India 2011). The huge number of industries in the early days along the bank line like jute mills, cotton mills, petrochemicals, engineering was paving to pay to increase the urban population. More than 90% population resides in urban areas (Census of India 2011); https://www.barrackpore.gov.in/. The climatic condition is a tropical pattern. The seasonal fluctuation in temperature shows the climatic variation. Summer seasons are hot with reach up 45° C and in winter, it reaches below 10 °C. The southwest monsoon brings rainfall  during the monsoon season (Das and Angadi 2020) (Fig. 1).

Fig. 1
figure 1

Location map of the study area

Full size image

As the research study is focused on 16 municipalities ‘micro level to quantify urban sprawl individuals in Barrackpore sub-division. The ULBs are represented by the name code as Kanchrapara (KNP), Halishahar (HSM), Naihati (NHT), Bhatpara (BHP), Garulia (GAR), North Barrackpore (N BKP), Barrackpore (BKP), Titagarh (TGH), Khardah (KDH), Panihati (PHT), Kamarhati (KMH), Baranagar (BN), South Dumdum (S DM), Dumdum (DM), North Dumdum (N DM) and New Barrackpore (NEW BKP). The details of Municipalities are mentioned in Table 1.

Table 1 Details of municipalities or ULBs in Barrackpore Sub-division
Full size table

Methods and materials

This piece of research work is the combination of LULCC detection cum landscape analysis and evaluating the urban growth potentiality for every small urban unit in the micro-scale by using the integrated geospatial model techniques. Several types of data are required for the selected domain measures which were obtained from different sources. The methods and materials have been divided into several sections as follows.

Data source

The availability of satellite data of MSS (Multi-Spectral Scanner), TM (Thematic Mapper), and ETM + (Enhanced Thematic Mapper) images for 1972, 2001 and 2016 have been obtained from the USGS earth explorer. The available geo-tiff format images having the projection system of UTM and reference datum is WGS 84. The acquired 60 m (MSS) and 30 m (TM/ETM) resolution of multispectral data were used for the thematic layer (LULC) production. The town maps of all municipalities have been collected from the different municipalities’ offices and geo-referenced the entire map by giving the projection systems (UTM/WGS 84) in ArcGIS software. The Survey of India (SOI) topo-sheet on a scale of 1: 50,000 and Google earth with a spatial resolution of 1 m have been used for accuracy assessment purposes (Table 2).

Table 2 Applied data and related information
Full size table

Pre-processing of data and classification

The acquired images in 1972, 2001 and 2016 with cloud-free for land use classifications were geometrically and radiometrically corrected first. The corrected spectral bands (optical) stacked to obtain multi-bands image for each considered year. The processed images were subset by the boundaries of each town as an area of interest (AOI). The clipped images were used as inputs for performed classification by supervised classification methods. The widely used maximum likelihood classifier (MLC) was applied for the classification (Sudhira et al. 2003; Li and Yeh 2004; Rawat and Kumar 2015). The supervised classification is a decision rule-based technique involved in taking sample area with known feature type in the ground and compared with the spectral signature of the pixel in the image (Lillesand and Keifer 1987). The MLC is a reliable and probability-based technique to apply to multi-temporal classification (Strahler 1980; Rawat and Kumar 2015). This study has used MLC because this algorithm can assemble the spectral signature of the pixel in a particular class and minimize the complexity (Li et al. 2014a, b), on the other hand, the main objective is to study LULC spatial change detections and expansion of urban landscape, not examine the internal feature structure, shape and dimension of built-up areas (Das and Angadi 2020). Several researchers have used the MLC algorithm for the classification (Sudhira et al. 2003; Mondal et al. 2016; Mithun et al. 2016). The signature of each class was tabulated and performed the classification. The LULC classification has been adapted from Anderson et al. 1976 classification system. The most accurate information about the land surface is a challenging task due to heterogeneous features generally occurs the mixed pixel problem in the image (Lu and Weng 2005). Mixed pixels arise in medium spatial resolution data (Jensen and Jungho 2007). In the urban lands includes heterogeneous features, such as buildings, trees, mixed vegetation, water, soil (Lu and Weng 2005; Dewan and Yamaguchi 2009). These problems can be addressed by taking ground GPS points, visual interpretation of local knowledge and using Google earth engine pro for the study area. The six LULC classes have been classified as final results. The statistics of land use for each town has been quantified for further analysis (1) vegetation, (2) agricultural land/cropland, (3) built-up land, (4) fallow land/open field, (5) water bodies, and (6) wetland.

Accuracy measurement of classified images was tabulated to test and check that the particular pixel has assigned to correct correspondence land features over the surface (Congalton 1991). The level of accuracy depends on the three parameters, i.e., level of classification, resolution of image and scale in the study (Rawat and Kumar 2015). A group of sample points has been taken from field GPS surveys and using the Google earth map system from all over 16 towns as a stratified random sampling to perform for accuracy assessment for the study. The minimum required sample in each category to calculate accuracy is 50 as suggested by (Manandhar et al. 2009). Hence, 100 samples per each class have been assigned in Erdas imagine v14 software to perform an accuracy assessment for each class by identifying the high spatial resolution from Google earth (1 m spatial resolution) and field points. The older year’s classification accuracy has been performed by the “historical view” of the google earth engine. A confusion matrix has been tabulated based on reference points and the user’s accuracy (error of commission) and producer’s accuracy (error of omission) were calculated for each class (Story and Congalton 1986; Congalton and Green 1999). Lastly, the overall accuracy and kappa coefficient for each were calculated and compared between 1972, 2001 and 2016 (Foody 2002, 2004).

Urban sprawl measures: landscape metrics and Shannon’s entropy model

The study has been attempted to analyze the process of the built-up growth of towns at Barrackpore Sub-division applying some selected landscape metrics. Although a wide variety of metrics were used to measure the patterns of urban growth by several researchers. The results of metrics depend on the resolution of satellite imagery, the accuracy level of image classification, selection of metrics and so on (Mithun et al. 2016). The different researchers have given a different opinion on the choice of metrics and confirm that many metrics are correlated and could produce redundant information (Ramachandra et al. 2012). In contrast, Shannon’s entropy is whispered to be a robust measure of the urban growth process. This method is preferable because it has marginal limitations, but not free from nuisances. Moreover, sometimes the contradictory relation would find in the result of both landscape metrics and Shannon’s entropy. Landscape metrics are unable to measure the degree of urban sprawl with black and white categorization (Bhatta and Giri 2012; Mithun et al. 2016). In the present study, Shannon’s entropy calculation was separately used to analyze urban growth for a proposed zoning approach to understand the performance of the selected metrics.

Landscape metrics

Landscape defines as a heterogeneous land area composed of a group of interrelating ecosystems patches that are repeated in a similar form (Forman and Godron 1986). Landscape metrics are an index that can be quantified in nature to describe the pattern and structures of the land (McGarigal and Marks 1995). Landscape metrics are known as spatial metrics through which they can comprehend and describe the causes as well as consequences of urban processes (Bharath et al. 2012). Several types of landscape metrics have been proposed and used to configure individual land class and whole land cover categories (Forman and Godron 1986; Frohn et al. 1996; O'Neill et al. 1999). The applications of landscape metrics include the landscape pattern analysis, biodiversity fragmentation, changes of landscape (Gardner et al. 1993; Dunn et al. 1991), relating landscape structures at different scales (Turner et al. 1989); and complexity of urban land structure (Herold et al. 2002). In the present study, the Fragstats program was used to calculate the landscape metrics (McGarigal and Marks 1994). Therefore, the variety of landscape metrics were derived from the estimation of urban structure for the small towns (ULBs).

  • Area and edge metrics This group of metrics quantifies the size of the patches and the amount of edge created by this patch in the landscape (McGarigal and Marks 1994, 1995). This section includes CA (class area), PLAND (percentage of land), LPI (largest patch index), MPS (mean patch size) and ED (edge density).

  • Shape metrics The patch shape and size influence the ecological processes. This group of metric measures the landscape configuration by calculating the shape complexity of patches. The shape metrics represent the regularities and irregularities characteristics of patches (McGarigal and Marks 1994, 1995). This segment includes AWMSI (area-weighted mean shape index).

  • Aggregation metrics Aggregation metrics quantify the tendency of patches to be spatially aggregated and also refers to as landscape texture. This metric indicates the dispersion, interspersion, sub-division and isolation in the landscape (McGarigal and Marks 1994, 1995). LSI (Landscape Shape Index), CLUMPY, IJI (Interspersion Juxtaposition Index), AI (Aggregation Index) is falling under these metrics.

Table 3 depicts the name of the metrics, formulas and description of the possible range of each landscape is being calculated. The different metric types were selected based on the potential utility to get information for different domains.

Table 3 Description of selected landscape metrics for urban class
Full size table

Shannon’s entropy model

In recent decades various studies applied Shannon’s entropy model to analyze and understand the equilibrium rate of the relative urban phenomenon at state, regional and country levels (Shannon 1948). The entropy is widely used to measure the degree of urban sprawl in a region with the integration of remote sensing and GIS approach and carried out with its spatial database (Das Chatterjee et al. 2016). Shannon’s entropy is an index or indicator which can measure the spatial concentration or dispersion in any spatial unit (Ao and Li 1998; Sudhira et al. 2004; Jat et al. 2008; Bhatta and Giri 2012; Mithun et al. 2016; Ramachandra et al. 2014a, b; Sahana et al. 2018). The structure of the model is calculated by using the formula below:

$${H}_{n}=-{\sum }_{i=1}^{n}{pi}^{*}{\mathrm{log}}_{e}({P}_{i})$$

where pi is the proportion of variable (built-up land) in the ith zone \(\left({P}_{i}=\frac{{x}_{i}}{{\sum }_{\mathrm{i}=1}^{n}{x}_{i}}\right)\), xi is the observed value of the variable in the ith zone and n is the total number of zones. The entropy value varies from 0 to log(n). The value closer to zero indicates the compact distribution and the value of near log(n) indicates the dispersed distribution. Higher values of entropy indicate sprawl (Bhatta and Giri 2012). The halfway mark of log(n) is measured as a threshold value, therefore if the entropy values are beyond the threshold can be called the sprawling city (Mithun et al. 2016). Relative entropy can be measured to scale up the entropy value into ranges from 0 to 1. The relative entropy (H′) for n number of zones can be calculated as (Thomas 1981):

$${H}_{n}^{^{\prime}}=\frac{{H}_{n}}{{\mathrm{log}}_{e}(n)}$$

The value 0.5 is considered as a threshold value. The value is higher than the threshold considered sprawl.

In our study, this model was used for evaluating the urban expansion in every ULBs (municipalities). The cities have their well-defined administrative areas. Organizationally these towns have a sub-division of municipal ward level, but these wards are not fixed in number and area, it is changeable with temporal manner. To overcome this problem, each ULB has been divided into four groups of North East (NE), North West (NW), South East (SE), and South West (SW) zones based on the administrative building point of the respective town. Each zone was further divided into a concentric circle pattern of 200-m radius buffer zones. Because the study aims to emphasize the urban growth in every spatial corner of these small towns. The urbanization process is not uniform and growth is defined in terms of directions (Ramachandra et al. 2014a, b). The questions regarding urban sprawl, i.e., how these small towns are facing rapid urban growth and how much built-up land is expanding through the time. It would be a detailed study on a micro-scale that will be planning oriented for planners to provide sustainable infrastructure.

Results and discussion

The spatio-temporal land use/cover changes in ULBs

LULCC of the municipalities from 1972 to 2016 is shown in the table and figure format (Table 4; Fig. 2). The output displayed that vegetation, agricultural land, wetland, water bodies have been declined in areas’ from 1972 to 2016; on the other hand, built-up areas and fallow land are increased during the decades in each town. The results of the overall accuracy of the classification of output images were more than 88% (1972), 93.3% (2001) and 95% (2016). The classified three-time interval images had a kappa index of 0.86, 0.92 and 0.94. The accuracy result indicated strong agreement of classified image and ground truth (Congalton 1991). The present study met the minimum requirement by overall accuracy and k statistics for LULC classification (Anderson et al. 1976). The classification results of 16 towns are described into four divisions; each of them discussed four towns in detail. The first section discussed KNP, HSM, NHT and BHP towns; the second section includes towns of GAR, N BKP, BKP and TGH; the third section contains ULBs of KDH, PHT, KMH and BN and lastly, four section is enclosed S DM N DM, DM and New BKP.

  • KNP, HSM, NHT and BHP

Table 4 ULBs wise land-use/land cover changes of classes (area in ha)
Full size table
Fig. 2
figure 2

Land use/cover map of municipalities: a KNP, b HSM, c NHT, d BHP e GAR, f N BKP, g BKP, h TGH, i KDH j PHT, k KMH, l BN, m S DM, n N DM, o DM p NEW BKP

Full size image

These four ULBs lie in the northern part of the sub-division. The KNP is a town where one of the railway workshops in India and jute mills are situated that bring the development and growth of the city. The decadal changes in LULC reveal that the built-up area has increased simultaneously. The vegetation cover has decreased from 53 to 36% (1972–2001) and 19% (2001–2016). Agricultural land and surface water bodies were reduced over time. The built-up land had gained 143 ha in 972–2001; and 126 ha in 2001–2016 intervals (Fig. 2a; Table 4a). Kanchrapara city contributes only to 3.8% of the residential area in the total sub-division area. The next city is HSM, known as the city of palaces. This town is facing the changing aspects of LULC. The area of Wetland was reduced by 13–1.7% from 1972 to 2016; and vegetation cover was also sharply decreased from 66 to 36% (1972–2001) and from 36 to 19% (2011–2016). Agricultural land had converted into open fallow land in some places of the town. Built-up areas have extended towards the eastern side rapidly. In 1972 built-up area was only 8.92 ha (1.3%), then increased to 108.37 ha (13%) in 2001 and 343.69 ha (43%) in 2016 of the town (Fig. 2b; Table 4b). NHT is one of the old cities. The colonial character has been established since the statute of the city. Changes in different land-use exposed that the city expanded in concreteness by 27–62% from 1972 to 2016. Urban sprawls are happing in this city because it encroached the adjoining Deulpara non-municipal (N.M) area over the period. The total spatial area of the city also has increased by 38%. The proportion of wetlands is gradually decreased in the area (Fig. 2c; Table 4c). BHP is industrial areas and the numbers of slums are more here because the workers from the outside made this area dense populated. It has an outgrowth extension according to the census of India 2011 report. Therefore, the area of the urban land is getting higher during a time interval. The growth rate of built-up land was 65% during 1972–1–2001 and 76% during 2001–2016. As the total area of the town were expanded due to merging with the villages and small census towns (namely ChakMulajor, Gurdaha, Madrail-Fingapara, Narayanpur and Sthitrapar), the statistics show that the other land use classes have added in area coverage during the 1972–2001(Fig. 2d; Table 4d). The urban sprawling is continued to notice in Bhatpara.

  • GAR, N BKP, BKP and TGH

These four municipalities are covered in the central part of the study area. Among all ULBs, the GAR municipality is small in size and less developed than others, but still, the built-up growth is continuing here. During the 1972–2001 time period, the built-up land was gain 100 ha (growth of 141%) and 60 ha (60% of growth rate) has added during 2001–2016. Now presently 73% of the land area is covered by built-up land. Local ponds and water bodies are going to be disappearing over time. The loss of surface water bodies was more than 45 ha area for 44 years (Fig. 2e; Table 4e). N BKP is one of the oldest municipal authorities in the study area. Based on the centralized rifle and metal gun factories, the settlements were growing up in this area. Even the older colonial aspects have been found in most places in the ULB. The urban built-up land was 105 ha in 1972; 312 ha in 2001; and 575 ha in 2016. The natural vegetated area and water bodies were dominantly declined from 51 to 26% and 17 to 8% of land throughout the 44 years (Fig. 2f; Table 4f). BKP is a developed small city that has many socioeconomic facilities with a densely populated area. The sub-divisional administrative activities have been done in this town only. It is a very fast-growing city. Mostly of amenities and facilities had played a role to attract people from adjacent small towns and villages. Every corner of town is full of built-up land. In 2016, 61% area was built up land where it was 12% of the land in 1971. Moreover, the non-built-up land was reduced to more than 50% of the land (Fig. 2g; Table 4g). TGH is mostly an industrial town dominated by many slum areas. It is small in size and compact full of residential areas. It had the highest population density as well as built-up density too. As the compactness of the built-up area has been still since 1972, the growth rate was not much higher. Increasing of built-up land was intruding the other non-built-up land, especially vegetation and surface water surfaces. It was an increase in built-up land from 41 to 54% and then to 58% during 1972–2001 and 2001–2016 (Fig. 2h; Table 4h).

  • KDH, PHT, KMH and BN

These ULBs are situated in the southern part of the area The KDH municipality had named the south Barrackpore municipality. In the urban centre, vegetation land, water bodies, wetland and agricultural land have been losing their land by 119 ha, 20 ha, 70 ha and 12 ha respectively; where built-up land and fallow areas were gained 248 ha and 3 ha during 1971–2016. Only 16% of the land was covered by built-up land in 1972, after that, it was increased by 47% in 2001 and lastly by a 64% increase in 2016. The ULB has added Keulia non-municipal village with a 10.9 ha area and expends the total spatial unit from 1972 to 2016 (Fig. 2i; Table 4i). PHT municipality was known as a trade and business center in the early days when the river route was the main source of communication. The name came from ‘Pannyahati’ (Emporium of the merchandise). These all factors had a role to grow this city densely. The total area of the city is 1867 ha within which 1141 (61%) is covered by built-up land in 2016. The growth rate of built-up land was 533% between 1972 and 2001 that indicates the fastest growth was taking place and 70% between 2001 and 2016. The urban land uses are spreading towards the eastern side of the city which can demolish the ecological balance in terms of loss of wetland, cropland and vegetation (Fig. 2j; Table 4j). In the next KMH urban centre is a total industrial area. This compact urban body shared 12% in 1972, 10% in 2001, and 7.5% of built-up land out of the total built-up land of the sub-division area. The growth rate of the built-up land was less because the settlement has been compact since before decades. The number of slums are huge here. LULCC dynamics reveal that vegetation and wetland were reduced 364 ha and 140 ha of land during 1972–2016 (Fig. 2k; Table 4k). BN municipality is very populated as well as a very denser residential area due to industrial influences and location near to the capital city, Kolkata. The growing residential settlement made this town a more compact urbanized area. The settlement growth process could destroy the natural ecological aspect in terms of reducing the number of water bodies and induced the area of wetland, vegetation and also agricultural land. The statistical evaluation shows that the total area of the ULB is about 890 ha. The Vegetation land was covered by a maximum area of 234 ha (30%) in 1972 and 681 ha (66%) by built-up land in 2016. No such area remains as a wetland. The small water bodies or urban ponds were filled to build the settlement (Fig. 2l; Table 4l).

  • S DM N DM, DM, and New BKP

These four ULBs are situated extreme southern part of the study area. The Dumdum zone is a highly developed area. It has three separate municipality boundaries. First, the S DM municipal has maximum urban growth due to its number of socioeconomic facilities for the citizens. The population density is very high over the periods. The built-up land was 18% in 1972; 61% in 2001 and 76% in 2016. It is noticeable how much rapid rate of impervious growth has happened in this town. The location near the capital city had major influences to grow and develop. Transportation and communication are very well connected. Metro rails; railways, the number of roads increased, near to the airport are the causes of fast development. Two villages (Garui and Natkal) have fused with the area of this town (Fig. 2m; Table 4m). Lastly, the N DM municipality has experienced more urban sprawl. The growth rate of urban built-up land over the period explained the settlements patches are increasing. As a result of compactness to neighboring South Dumdum, Dumdum area and near Kolkata city people construct the settlement in these areas. The development and growth would support this place to get broader areas. Near villages (Bandra, Bisarapara, Fattullapur, Finga and part of Sultanpur) were merged up with the town. The total spatial unit has an extent of 920 ha. During 1972–2001, the built-up area has expansions of 446 ha from 2001 to 2016, it gained 671 ha (Fig. 2n; Table 4n). The DM urban centre is a small town, but the growth of the built-up land was very high because it was fused with part of Sultanpur village and extended the spatial unit. The present conditions of the town are compact in the built-up area. More than 72% area was dominated by the impervious areas in 2016. The other non-built-up land had decreased by the area (Fig. 2o; Table 4o). New BKP was a newly built town among the sixteen ULBs, so the number of population, as well as built-up density is less compared to others. But this ULB is facing the fastest built-up land as it is part of the urban agglomeration process of KMDA. In 1972 the area distribution of land use classes was 58%, 18%, 2.05%, 5.7%, 0.12% and 13% of the vegetation, agricultural land, built-up land, wetland, fallow land and water bodies; it has changed the statistics for different land uses in 2001, i.e., 56%, 5%, 23%, 3.4%, 11% and 7% of correspondence land classes; and it was declined to 16%, 7%, 0.04%, 5% for vegetation, agricultural land, wetland and water bodies but increased to 66%, 17% for built-up and fallow land in 2016 (Fig. 2p; Table 4p). The New Barrackpore ULB spatial unit has merged with two adjacent villages (Agapur, Kodalia villages) and expanded it.

Fig. 3
figure 3figure 3

Class level landscape metrics for municipalities: a CA, b PLAND, c NP, d PD, e LPI, f ED, g MPS, h AWMSI, i LSI j CLUMPY, k IJI, l AI

Full size image

Urban expansion (sprawl) analysis

The spatial pattern of urban sprawl by landscape metrics in ULBs

The selected metrics compute the characteristics of entire spatial landscapes in all ULBs. These can quantify the proportion of the landscape in individual class based on fragmentation, shape, edge and contagion. The urban structure with spatial patterns examined depending on urban patch types, patch density, patch diversity and patch evenness. The complexity, compactness, aggregation, dispersions and sprawling characteristics would measure by these indices in all municipalities that would be suitable to take planning decisions for urban planners and local government in terms of environmental sustainability. The landscape metrics were calculated by using the binary image in Fragstats software. The scores of metrics would justify describing the landscape pattern (Fig. 3).

  • Area of urban land (CA)

The class area (CA) is the sum of the areas of all patches in the corresponding land class (McGarigaland Marks 1994). The urban class patches have been calculated by these metrics (Table 5). Figure 3a illustrated that the higher urban area (in hectares) expanded in BN (306.45 ha), BHP (257.52 ha), KMH (468.81 ha), N DM (446.59 ha), PHT (564.47 ha) and S DM (661.66 ha) municipalities during 1972–2001; wherein BHP (494.48 ha), HSM (243.70 ha), NHT (212.43 ha), N DM (671.47 ha) and PHT (471.31 ha) ULBs during 2001–2016.

  • Percentage of land (PLAND)

Table 5 Different landscape metrics of urban land use in ULBs
Full size table

PLAND is the proportion of a particular class in a landscape. The temporal PLAND indicates that the higher growth trend in built-up class over the periods in each municipality. The higher percentage growth of the built-up class has met in BN (34%), BKP (33.53%), DM (27.82%), GAR (26%), KMH (37.62%), KDH (34.57%), PHT (31.88%) and S DM (41.66%) ULBs from 1972 to 2001; where HSM (29.99%), New BKP (42.07%), N DM (30.47%) and PHT (25.30%) ULBs from 2001 to 2016. The built-up percentage increases in the landscape indicate to understand the sprawling characteristics of municipalities (Fig. 3b; Table 5).

  • Number of patches (NP)

NP is the total number of built-up patches in a given landscape. It is an indicator to show the level of fragmentation in a specific class in the landscape. The result shows that the number of built-up patches has increased in all ULBs during 1972–2001. In that period, the built-up patches had moved from the city center to the fringe or peripheral areas and exhibits fragmented urban growth in those places. But while in 2016 the reduced number of patches indicates that the merging of patches and form into a single compact patch (Fig. 3c). BHP municipality only has experienced an increase in the number of patches over the periods (Table 5).

  • Patch density (PD)

PD is an indicator of urban fragmentation (Ramachandra et al. 2014a, b). Patch density increases as the number of patches increases. A similar trend showed that in 2001 the fragmentation of built-up patches was increased and reduced in 2016. In 2001, the urban patches were gradually increased from the CBD of towns that specifies that the fragmentation of the landscape and it indicates the sprawl in a town. Reducing of patch density indicates that less number of patches because those had combined with the small patches and form a single larger patch unit (Fig. 3d; Table 5). HSM, New BKP and N DM had a higher patch density in 2001. The gradual fragmentation of the urban landscape happened in the BHP municipality over the periods. The completely occupied built-up class with low patch density in the core area, as well as peripheral areas of each ULBs in 2016, indicates the compact urbanization in city centers and the outskirts of sprawl. The Barrackpore sub-divisional all municipalities areas having a sprawling effect.

  • Largest patch index (LPI)

LPI equals the area of the largest patch of the corresponding patch type divided by total landscape area, multiplied by 100 (McGarigaland Marks1994). The LPI is widely used as an indicator of landscape fragmentation (Sun et al. 2014; Zhang et al. 2004, 2012). The index has increased year by year. Figure 3e showed that the largest built-up patches were found in BN, DM, BKP, GAR, KMH, KDH, PHT and S DM municipalities in 2001. While in 2016 the maximum value of LPI was in N BKP, New BKP and N DM areas (Table 5; Fig. 3e). This metric helps to identify the growth pole conditions of a town in different years (Ramachandra et al. 2014a, b).

  • Edge density (ED)

ED is also an indicator to measure the fragmentation and the spatial heterogeneity of the landscape (Wei and Zongyi 2012). In the study area during 1972–2001, the value of edge density was higher among the all ULBs that compute the level of fragmentation in built-up was more and indicates the sprawl. While in 2016 the compact or clumped urban growth had experiences in BKP, GAR, KMH, KDH, PHT, S DM and TGH municipalities with relatively low edge values and dispersed urban growth was found in BHP, HSM, KNP, New BKP, N BKP and N DM areas with relatively higher values (Fig. 3f; Table 5).

  • Mean patch size (MPS)

MPS is the sub-division measures in class or landscape metrics. Mean patch size quantifies the sum of the areas of all patches of the corresponding patch type divided by the number of the same type and convert to hectares (McGarigal and Marks 1994, 1995). There is an inverse relationship between MPS of built-up class and degree of fragmentation. The lower value of the index represents the greater fragmentation while the higher value reveals the aggregated growth of the city. Here, in 2001, the value index was very low among the ULBs that indicate the maximum fragmentation in built-up class and sprawling was happening in the period; but in 2016 the high index value reflects the larger patches due to compactness. Presently the ULBs are becoming compact (Fig. 3g; Table 5).

  • Area weighted mean shape index (AWMSI)

The weightage is higher in larger patches and lowers in smaller patches. The metric is useful to analyses the structure of the landscape through the spatial scales by calculating the complexity of urban patches based on their size (Huang et al. 2009). The index represents shape irregularities in the patches. The range smaller to higher value indicates regular to the irregular shape with complexity. In 1972, the AWMSI value was low that represents the compactness of CBD areas in every town; while in 2001 as the value increases indicates the fragmentation of built-up land increases over the ULBs. However, in 2016, it is noticeable that some ULBs had lower values than 2001 time period that indicates the compactness was happening due to the growth and development, while some ULBs had increased the value means fragmented the urban landscape and keep continuing the growth of sprawl in these towns; namely, HSM, New BKP and N DM (Fig. 3h; Table 5).

  • Landscape shape index (LSI)

LSI measures the complexity of the patches. The value closer to zero represent compact of urban patches and higher value indicates the desegregation of the patches (Table 5). The LSI quantifies with the lower values in all ULBs in 1972 indicates the compactness of the towns, but the value has increased in 2001 indicates the fragmentation of the landscape and desegregation in all municipalities. However, according to this metric, the trend of sprawl has shown at city outskirts along with the centers of the city due to the decline of value in 2016 excluding Bhatpara and Halishahar municipality (Fig. 3i).

  • CLUMPY

CLUPMY estimates the aggregation of urban patches. The range of CLUMPY is ‘− 1 < 0 > 1’. Disaggregation of the patches are denoted by value ‘− 1’, the random distribution of patch is represented by value ‘0’ and ‘1’ indicates the patch distribution is aggregated. In this study, the ULBs were more aggregated in 1972 with value more than about 0.7; whereas the value was decreased (0.4 − 0.65) in 2001 and 2016 that indicates the desegregation or fragmentation of patches at outskirt (Fig. 3j; Table 5).

  • Interspersion and juxtaposition index (IJI)

The IJI method calculates the interspersed of patch types. Higher values indicate the urban patch types are maximally interspersed and juxtaposed to each other’s (equally adjacent to each other) and lower values approached the patch types are poorly interspersed. The range varies from the index 0–100. The result (Fig. 3k) also carried out that there has been a growth of built-up parches in 2001and 2016 in peripheral areas (Table 5).

  • Aggregation index (AI)

This index measures the aggregation of the town. The higher value of AI is maximally aggregated and the lower values equal to the maximal desegregated. Figure 3l depicts the fact that the urban patches were compact or aggregated in nature in 1972, but the fragmentation was occurring in the landscape with lower values and implies the desegregation of urban patches in the ULBs in 2001. And lastly, in 2016, the small fragmented patches in municipalities peripheral were getting merged up and combined with larger patches that indicate compactness or aggregation of the urban landscape. Therefore, the AI value has increased in 2016. Over the periods, every small town and ULBs have tremendously faced urban growth at centers as well as in peripheral or a fringe region (Table 5).

Shannon’s entropy for urban sprawl analysis

The area of built-up area in all municipalities of every buffer zone for each temporal image has been calculated by clipping off the classified vector format. This index is widely used because it can identify both urban sprawl and growth. The results of the Shannon entropy index are presented in table and figure, respectively. The continuously increasing trend with time has been noticed in built-up land at each small ULB (Figs. 4 and 5; Table 6). The elaboration and illustration have been discussed by four groups of ULBs.

  • KNP, HSM, NHT and BHP

Fig. 4
figure 4

Zone wise built-up land in 1972, 2001, and 2016 at ULBs: a KNP, b HSM, c NHT, d BHP, e GAR, f N BKP, g BKP, h TGH, i KDH j PHT, k KMH, l BN, m S DM, n N DM, o DM, p NEW BKP

Full size image
Fig. 5
figure 5

Zone wise Shannon’s Entropy index in 1972, 2001 and 2016 at ULBs: a KNP, b HSM, c NHT, d BHP, e GAR, f N BKP, g BKP, h TGH, i KDH, j PHT, k KMH, l BN, m SDM, n N DM, o DM, p NEW BKP

Full size image
Table 6 Results of the Shannon entropy in ULBs
Full size table

The results of entropy models in KNP, HSM, NHT and BHP cities reveal that the NW direction was faced with the more urban landscape in KNP town. The log (n) was 0.95, 1.0, 1.08 and 1.18 in NE, NW, SE and SW Zone. All entropy values were above the half of the threshold level that indicates sprawl happens over the region. The growth rate of sprawl was from 1972 to 2001 by 59% in NE, 46% in NW, 21% in SE and 12% in SW while during 2001–2016 the growth rate has decreased sharply (Figs. 4a and 5a; Table 6a). The scenario of urban sprawl was rapid in the HSM town. The urban sprawl is noticed in the NE and SE zones. In 1972, the entropy value was below the threshold level in the NE and SE zone due to compact urban land patches; whereas the 140% and 249% growth in entropy in these zones from 2001 to 2016 (Figs. 4b and 5b; Table 6b). The NHT is becoming a compact town in terms of built-up land. The NE zone and SE zone are kept continues (dispersed) to grow the settlement. However, in the NW and SW zone are aggregated. In 2016, the entropy value reveals that the NW zone had negative growth entropy growth and the SW zone had negative growth in 2001 and very slow growth in 2016 (Fig. 4c and 5c; Table 6c). BHP is one of the Towns which have a settlement patch growth constantly. All zones over the temporal years had an entropy range above the threshold mark. The log (n) for four zones were 1.38 (NE), 1.0 (NW), 1.28 (SE) and 1.28 (SW) (Figs. 4d and 5d; Table 6d).

  • GAR, N BKP, BKP and TGH

The southern part was affected by urban sprawling in GAR urban centre. Due to the small size of the town maximum extension was happens during 1972–2001. In the NE zone, the entropy values were 0.75, 0.78 and 0.77 for the year of 1972, 2001 and 2016. NW and SW zone was more compact due to the spatial limitation in the western side along with River side. Lastly, in the SE zone had an entropy value with a growth rate of 4.35% in, between 1972 and 2001 and decreased from 2001 to 2016 with, 2.6%. Decreased of entropy values indicate the compact or aggregation of urban land use growth over the spatial unit (Figs. 4e and 5e; Table 6e). The central part of the N BKP municipality is covered by the Air force basement area. Hence, the NE, SE and lower part of the SW zone are residential prone and it’s constantly growing. Table 6f shows that during 2001–2016 the NE zone became compact where SE and SW zone are dominantly sprawling areas (Figs. 4f and 5f). The BKP town is situated in the central part of the study area. Here, the town has urban sprawling effects on the eastern side. Due to the dense conditions of the eastern side, now the western side is starting to grow built-up land. The result of entropy revealed that all sectors were above the halfway of log (n) threshold level (Table 6g). The Sharpe decline of the growth rate of entropy explained the aggregation of the urban landscape (Figs. 4g and 5g). The TGH urban centre is full of industries and dense settlement. However, urban sprawling is less over the period because of this the town is in less spatial size. Only on the eastern side has to spread in built-up classes. The entropy value indicates the growth was negative in the NW and SE zone during 2001–2016 (Figs. 4h and 5h; Table 6h).

  • KDH, PHT, KMH and BN

The urban sprawl index for these four ULBs showed that the KDH town has grown on the western side. All zone had an entropy value more than the threshold level. During 1972–2001, the growth rate of entropy was 52.06%, 1.67%, 2.41% and 4.29% for NE, NW, SE and SW zone due to dispersions of urban patches. Although it became compact or aggregated due to the decline in the growth rate of entropy during 2001–2016 (Figs. 4i and 5i; Table 6i). The spatial area is larger in the PHT municipality. So the settlement had settled in the peripheral areas of the town over the periods and made this area full of settlement. The result of sprawl measurement evaluated that in the NE zone the value was 0.98, 1.16 and 1.22 {log (n) = 1.36} in 1972, 2001 and 2016; 0.85, 0.91 and 0.92 {log (n) = 1.08} in NW zone; 1.09, 1.27 and 1.26 {log (n) = 1.36} in SE zone; 0.89, 0.97 and 0.97 {log (n) = 1.04} in SW zone of the town (Figs. 4j and 5j; Table 6j). The KMH industrial area has been quite compact since 1972. But still, the town is facing urbanization. The sprawl rate was higher in every zone during 1972–2001 that indicates the dispersion or fragmentation of landscape-level in the town. But the rate of growth of entropy was less during 2001–2016 which specifies the near to aggregation or compact growth (Figs. 4k and 5k; Table 6k). Baranagar town (BN) is a highly populated and dense area. The increasing trend of urban sprawl has observed during 1972–2001 through increases in the entropy value. The decadal growth rate of entropy during 1972–2001 was 22.01%, 2.93%, 5.57% and 3.90% in NE, NW, SE and SW zone. While it has decreased its value to 2.1%, 1.4%, 0.5% and 0.27% in the respective zone between the years of 2001–2016 (Figs. 4l and 5l; Table 6l).

  • S DM, N DM, DM and New BKP

These four ULBs are much more compact in the present. The spatial urban sprawl index for these towns revealed that there was a growth of built-up land during the concerning time in SDM town. The entropy value illustrated an increase in entropy from 1972 to 2001 at all zones. It has changed during 2001–2016 and declines the values that quantify the combined small built-up patches into a big urban patch. The character of the city now becomes compact (Figs. 4m and 5m; Table 6m). The N DM town had experienced a higher sprawl impact. A significant increase in entropy is observed in four zones during 1972–2001 to 67.4%, 21.4%, 27.7% and 44.4% rate of growth. It is noticeable that concerning the next entropy values have declined and converted to compact municipal land (Figs. 4n and 5n; Table 6n). The Dumdum municipality is also a very denser built-up area. The zone-wise entropy values exhibited that the entropy value has increased from 0.91 to 1.03 in between 1972 and 2001; and has decreased the next period in the NE zone. The result of dispersion has noticed in the first period. The reasons to decline the entropy due to built-up fill in every corner of the town (Figs. 4o and 5o; Table 6o). The N BKP urban centre has recently built the town and due to overflow population pressure from the adjacent municipality, it has experienced the rapid growth of settlement and built-up land. In 1972, every zone had less built-up land. As time goes up, the built-up land grew up in this town. The entropy value showed that the sprawl happens during 1972–2001 and in 2016 it would become a compact settlement area due to the agglomeration process (Figs. 4p and 5p; Table 6p).

Conclusion

The present research emphasized on town level studied to measure the urban growth dynamics. It considered the municipal administrative boundaries of the Barrackpore sub-division to study the LULCC phenomenon and examine the degree of urbanization at the micro-scale. Moreover, these small towns have data on socioeconomic variables. The urban union contains the diversity of land use characteristics include green vegetation, water surfaces and open space as well. In this study, the dynamics of urban expansion of 16 municipalities have been studied for the period between 1972–2001 and 2001–2016 using landscape metrics and Shannon’s entropy index incorporated with remotely sensed data and GIS techniques. The key findings show that the degree and scale of the built-up area have increased expressively over time; where, the vegetation land, Agricultural land and water bodies have significantly declined, which foremost, put a negative impact on the environment in terms of ecology, net productivity and changes of urban micro-climate (Das and Angadi 2020). The presentation of selected metrics was verified for the municipalities and discussed relatively. Different municipalities had different urban growth characters due to their heterogeneity activities. Industrial economy-based towns are dense since the first period. The results of CA indicate that the built-up area is increasing in all ULBs with time. The PLAND analysis demonstrated a similar trend. The outcome of NP and PD analysis illustrated that the urban landscape fragmentation and sprawling during 1972–2001 but declining the number indicates the compactness of the city during 2001–2016. LPI measurement signifies the rate of fragmentation. From 1972 to 2001, most of the municipality had increased the value of LPI excluding BHP, HSM, NEW BKP, N BKP and N Dm, etc.; while LPI was maximum in these ULBs during 2001–2016. ED analysis demonstrated that the sprawling character of the municipalities. MPS and AWMSI indicate the same trend of measurement. Lastly, the four aggregation metrics were applied to analyze municipal aggregation and desegregation in municipalities. All the metrics were performed well. Among the 16 ULBs, particularly, Halishahar, Bhatpara, New Barrackpore, North Dumdum municipality continues to sprawl in the respective landscape. The result of landscape metrics in different ULBs showed that sprawl has happened during the 1972–2001 periods and later these small municipalities had become compact due to more demand for land and population pressures. The small urban class patches at the peripheral zone combined with a single larger patch concerning time. The Shannon’s entropy was applied for urban sprawling measurement for ULBs individually with four directions of zones. The result of entropy illustrated that direction and zone wise sprawl of urban land. Mostly eastward direction of towns has a probability to increase in the future while the western side is congested and compact. The index value has provided visualization and quantifying the expanding urban footprint in the study area. Hence, the findings of the study would help to take management and planning for the environmental aspects of the towns. The government needs to do supervision in sustainable planning for the municipalities with good physical shape and sustenance of natural resources. As the current study highlighted that the loss of vegetation, agricultural land and water bodies happened in all small towns so that there would be important to take an integrated approach to protect the natural resources and to ensure the well being of people’s livelihood.