**Recent Progress in Urbanisation Dynamics Research**

Editors

**Iwona Cie´slak Andrzej Biłozor Luca Salvati**

MDPI • Basel • Beijing • Wuhan • Barcelona • Belgrade • Manchester • Tokyo • Cluj • Tianjin


*Editorial Office* MDPI St. Alban-Anlage 66 4052 Basel, Switzerland

This is a reprint of articles from the Special Issue published online in the open access journal *Land* (ISSN 2073-445X) (available at: www.mdpi.com/journal/land/special issues/Urbanisation Dynamics).

For citation purposes, cite each article independently as indicated on the article page online and as indicated below:

LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. *Journal Name* **Year**, *Volume Number*, Page Range.

**ISBN 978-3-0365-3023-9 (Hbk) ISBN 978-3-0365-3022-2 (PDF)**

© 2022 by the authors. Articles in this book are Open Access and distributed under the Creative Commons Attribution (CC BY) license, which allows users to download, copy and build upon published articles, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications.

The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons license CC BY-NC-ND.

## **Contents**



## **About the Editors**

#### **Iwona Cie´slak**

Iwona Cieslak is a professor at the Department of Socio-Economic Geography at the Institute ´ of Spatial Management and Geography, Faculty of Geoengineering, University of Warmia and Mazury in Olsztyn (Poland). Her research interests are mainly focused on the fundamental aspects of spatial analysis; GIS; and urbanism. These activities concern the development of methods for investigating the phenomena of city planning and sustainable development as well as evaluation of spatial conflicts. She is the author of more than 70 scientific papers on the subject. She leads and participates in scientific projects. She has also been an academic teacher for many years, teaching in this field.

#### **Andrzej Biłozor**

Andrzej Biłozor is a professor at the Department of Socio-Economic Geography at the Institute of Spatial Management and Geography, Faculty of Geoengineering, University of Warmia and Mazury in Olsztyn (Poland). His research interests are mainly focused on spatial planning; urban analysis; optimization of urban space; spatiotemporal analysis; fuzzy set theory; changes in land use and land cover. These activities concern the development of methods for investigating the phenomena of city planning and sustainable development, evaluation of spatial conflicts and optimization of spatial processes. He is the author of the 76 scientific papers on the subject. He leads and participates in scientific projects. He has also been an academic teacher for many years, teaching in this field.

#### **Luca Salvati**

Luca Salvati, M.Sc. Statistics and Economics, Sapienza University of Rome, PhD at the Faculty of Economics of the same university, is a researcher (Class B, holding a full professor habilitation in economic statistics) at the Department of Economics and Law of the University of Macerata. Since 2001, he has been staff researcher at the National Council for Research in Agriculture and the Analysis of Agricultural Economics (CREA). From 2006 to 2010, he was permanent researcher at the National Institute of Statistics (ISTAT). He has carried out continuous research periods abroad (2012-2014) at various public universities and research institutions in France (Toulouse), Spain (Barcelona) and Greece (Athens). He is an expert in the topics of official statistics, spatial statistics, Geographic Information Systems and Remote sensing applied to socioeconomic and environmental issues. He has conducted studies on topics of economic statistics, urban economics, regional demography and sustainable development, using exploratory multivariate statistics and geographic information systems for decision support, more recently developing studies on the monetary evaluation of natural resources. He has held courses in Economic Statistics (University of Rome Tor Vergata), Basic Mathematics (University of Rome La Sapienza), Multivariate Statistics (University of Roma Tre), Regional Economics (University of Camerino, University of Eastern Piedmont), and Strategic Evaluation of environmental impact (University of Roma Tre). He is a member of the teaching staff of the doctorate of for economics and finance, curriculum of Economic Statistics (University of Rome La Sapienza). He has supervised master's and doctoral theses, both in Italian universities and abroad. He has published more than 30 printed books and over 500 scientific publications.

## **Preface to "Recent Progress in Urbanisation Dynamics Research"**

This book is devoted to in depth analysis of past, present and future urbanization processes all over the world. New methods and assessment techniques were also investigated extensively. The development of science and technology has provided many new tools for the observation of urbanization processes and the formulation of conclusions about this phenomenon. Recent Progress in Urbanization Dynamics Research contains a broad range of papers presenting the complexity of the urbanization process, and varied approaches focusing on different aspects of urban development in relation to land development. This issue of *Land* is dedicated to the dynamics of urbanization processes, rapid changes in urbanization and scattered development of urban areas, zoning, sustainable development, urban sprawl, spatial conflicts, the real estate market, transport accessibility, spatial analyses involving GIS tools, and map-making methods.

> **Iwona Cie´slak, Andrzej Biłozor, Luca Salvati** *Editors*

## *Editorial* **Land as a Basis for Recent Progress in the Study of Urbanization Dynamics**

**Iwona Cie´slak 1,\* , Andrzej Biłozor <sup>1</sup> and Luca Salvati <sup>2</sup>**


Urbanization is one of the most dynamic processes occurring on the Earth. The urban population continues to increase due to the benefits of the urban lifestyle and the economic and social aspects of urbanization. This process became especially intensified in the second half of the 20th century, when the number of urban dwellers rose from 751 million to 4.2 billion in 2018. According to international forecasts, by 2050, the world's urban population will grow by another 2.5 billion, i.e., by 68%. However, the rapid rate of urbanization leads to serious environmental, spatial and socio-economic problems such as soil degradation, loss of urban ecosystem services, urban heat islands and air pollution. Health problems, urban poverty, rising crime and overcrowding are also becoming very acute. However, it should be noted that urbanization is perceived positively in many dimensions. Planned and sustainable urban development is the basis of a properly functioning economy, promoting a rise in living standards through higher quality of education and improved access to healthcare, culture and art. Researchers thus devote much attention to this process and monitor it constantly. Until recently, statistical data, including population or investment growth in administrative units, were the primary source of information for studying urbanization processes in the global or local dimension. This process could not be followed with a frequency suited to the rate of its progress based on data on spatial transformation. However, enormous databases containing a broad spectrum of spatial information, as well as new and rapid tools for processing spatial data have been available to researchers for more than 20 years, not only locally, but also on the international and global scale. The scope and quality of research on urbanization processes based on land data are also increasing. According to estimates, in 1990–2000, the average annual increase in built-up land was approximately 3.6% in developing countries and only 2.9% in industrialized countries. The most dynamic changes in land use towards urbanization have been observed in East Asia, including the Pacific region, and South-East Asia, where the increase in urbanized land reached 7.2% and 6.4%, respectively. In Europe, the annual increase in urban area does not exceed 2% in the most rapidly developing areas and is close to zero in rural areas.

Land use (LU) studies have become easier thanks to land cover (LC) observations. Land cover analyses are becoming increasingly advanced and provide knowledge on all dimensions of urbanization at various scales of reference. Most of the world's urban areas have experienced significant changes in land cover over the years. Land data, generally referred to as Spatial Information System (SIS) data, are increasingly used in research. The SIS provides information on LU/LC and currently covers various areas of interest, including administrative boundaries, transport and hydrographic networks, terrain, settlement and anthropogenic structures. These data are expanded to include environmental and social dimensions that describe not only the changes taking place in the environment, but also in the quality of life. SIS databases are being created at all levels of detail. Various types of national spatial information systems provide access to key data. In this case, spatial data

**Citation:** Cie´slak, I.; Biłozor, A.; Salvati, L. Land as a Basis for Recent Progress in the Study of Urbanization Dynamics. *Land* **2022**, *11*, 118. https://doi.org/10.3390/land11010118

Received: 4 January 2022 Accepted: 10 January 2022 Published: 12 January 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

refer to location (coordinates in the adopted reference system), the geometric properties and spatial relationships of objects that are identified in relation to the Earth and can be used in the analyses of urbanization processes. Spatial information can be presented cartographically and analyzed to identify urbanization patterns. Based on this reasoning, the Earth is one of the primary sources of information on urbanization processes.

Urbanization causes deep changes in space and gradually modifies the land use structure. There is an enormous need for monitoring urban space as well as changes that occur in areas directly subjected to urbanization pressure. The development of science and technology has provided many new tools for the observation of urbanization processes and the formulation of conclusions about this phenomenon. *Recent Progress in Urbanization Dynamics Research* contains a broad range of papers presenting the complexity of the urbanization process, and varied approaches focusing on different aspects of urban development in relation to land development emphasize the need for further research on urbanization. Different chapters are dedicated to the dynamics of urbanization processes, rapid changes in urbanization and scattered development of urban areas, zoning, sustainable development, urban sprawl, spatial conflicts, the real estate market, transport accessibility, spatial analyses involving GIS tools and map-making methods. Specifically, the authors focus on:


All of the discussed topics relate to urbanization processes that are observed on a daily basis as well as the methods and techniques for monitoring these phenomena. The relationship between urbanization and space seems obvious, but it is also interesting and multi-layered. Scientific and technological advancement has generated numerous tools for observing urbanization processes, conducting analyses and formulating conclusions about urbanization. *Recent Progress in Urbanization Dynamics Research* presents the latest research trends by relying on the extensive body of knowledge relating to urban geography. Special emphasis was placed on the global effects of urbanization, and this multidisciplinary phenomenon was analyzed with the use of satellite and photogrammetric observation techniques and GIS tools.

**Author Contributions:** Conceptualization, I.C. and A.B.; writing—original draft preparation, I.C. and A.B.; writing—review and editing, I.C., A.B. and L.S. All authors have read and agreed to the published version of the manuscript.

**Acknowledgments:** We thank all the reviewers for their feedback on earlier versions of the manuscripts in this Special Issue.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Review* **Review of Experience in Recent Studies on the Dynamics of Land Urbanisation**

**Andrzej Biłozor and Iwona Cie´slak \***

Department of Socio-Economic Geography, Faculty of Geoengineering, University of Warmia and Mazury in Olsztyn, Prawoche ´nskiego 15, 10-720 Olsztyn, Poland; abilozor@uwm.edu.pl

**\*** Correspondence: isidor@uwm.edu.pl

**Abstract:** Urbanisation rapidly accelerated in the 20th century. Along with the increasing dynamics of this phenomenon, the desire to know its origins and its course as well as to anticipate its effects is also growing. Investigations into the mechanisms governing urbanisation have become the subject of numerous studies and research projects. In addition, there has been a rapid increase in the number of tools and methods used to track and measure this phenomenon. However, new methods are still being sought to identify changes in space caused by urbanisation. Some of the indicators of urbanisation processes taking place include quantitative, qualitative and structural changes in land use, occurring at a certain time and place. These processes, related to human activity at a given time and in a given area, are determined by spatial diffusion, usually spreading from the city center towards the peripheral zones. Changes in land use involve the transition from less intensive to more intensive forms of land use. The constant effort to acquire new land for development, the search for alternative solutions for the location of investments and the need to determine the correct direction of development generates the need to constantly apply newer methods in the study of the dynamics of urbanisation processes. This paper presents an overview of recent studies and the most interesting—in the authors' opinion—methods used in research into the dynamics of urbanisation processes. The main objective of the authors was to produce a compendium to guide the reader through the wide range of topics and to provide inspiration for their own research.

**Keywords:** urbanisation; land use; dynamics of urbanisation; methods for mapping

#### **1. Introduction**

Urbanisation is one of the most visible manifestations of global change occurring for anthropogenic reasons [1]. It can be defined as a process of population concentration, mainly in urban areas, which also determines the growth of the urban population and its share in the population of a given area. Urbanisation is described as a cultural and civilizational process reflected in the development of cities, the increase in their number, the expansion of their area, the progressive concentration of the population in their immediate vicinity, the expansion of non-agricultural sources of income, the acceptance and assimilation of urban standards, customs, etc. It is a complex and diversified process, taking place with differing intensity and speed and with different effects in various countries and regions of the world [2].

Urbanisation is both a process and a state [3]. As a process, it involves the changes in human activities and socio-economic behavior described above, etc. As a state, i.e., the result of a process, urbanisation most often refers to the number and size of urban settlement units, the proportion of the urban population in relation to the total population of the country or region (urbanisation coefficient), the concentration of the population in large cities or its dispersion, and the area occupied by city buildings [4]. These two dimensions are interrelated and are dependent on each other. It should also be noted that urbanisation is a global phenomenon, which is irreversible and inherent within human development. It is associated with the scientific and technological revolution, with the

**Citation:** Biłozor, A.; Cie´slak, I. Review of Experience in Recent Studies on the Dynamics of Land Urbanisation. *Land* **2021**, *10*, 1117. https://doi.org/10.3390/ land10111117

Academic Editor: Dagmar Haase

Received: 15 September 2021 Accepted: 18 October 2021 Published: 21 October 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

concentration of productive forces and forms of social relations, with the spread of the urban style of life, with changes in social relations and ties, with the transformation of the rural population into diversified non-agricultural groups and with the modernization of the entire settlement network [5,6].

Nowadays, urbanisation is approached as a multidimensional complex of economic, social, demographic, cultural and spatial events and phenomena leading to the growth of cities and an increase in the share of the urban population, the population concentration in spatially expanding urban areas, the concentration of economic and administrative activities generating an increase in the importance of cities, shaping specific cultural patterns of urban lifestyle and specific arrangements of landscape and architecture [7].

The study of urbanisation processes can be based on each of these dimensions. The demographic dimension of urbanisation is related to the increase in the number of city residents, caused both by rural population migration to cities and by natural growth in cities and the creation of new settlement units. With regard to rural areas, demographic urbanisation, especially in the suburban zones of highly developed countries, is related to the population outflow from cities to the countryside. The social dimension of urbanisation is expressed in the spread and deepening of the "urban lifestyle", i.e., in attitudes, skills, behavioral patterns and even personality traits that are characteristic of metropolitan communities, both in the city area and in the countryside [8–10]. The economic dimension of urbanisation involves changes in occupational structure and employment. The occupational diversification of the population is becoming more pronounced in the city. The number of people employed in various services is growing at the expense of agriculture and crafts. The proportion of employment in traditional agricultural jobs is changing in rural areas in favor of non-farm jobs. All of the above dimensions of urbanisation are linked by spatial changes in the appearance and organization of the surrounding areas. Therefore, the spatial dimension of this phenomenon is one of the most interesting topics for research into the dynamics of urbanisation, since it provides a basis for inferring the causes and effects of the aforementioned aspects. The spatial dimension of urbanisation applies to both urban and rural areas. Spatial urbanisation is understood as the expansion of the urban landscape [11–15] and includes an increase in the urban investment area as well as the saturation of villages with urban-like infrastructure and buildings. In rural areas, spatial changes are associated with transformations in land use. The area of land used for farming is decreasing in favor of other, non-agricultural forms of use. Changes are taking place in the physiognomy of the buildings and the village morphology, as well as in the modernization of the technical infrastructure.

Current research primarily focuses on demographic and spatial urbanisation. It should be stressed that these are essentially just objectively separated manifestations of the same phenomenon. However, no spatial effects of urbanisation exist without demographic changes. The anthropogenic background of this phenomenon makes the demographic aspect an overriding factor in understanding the causes of spatial change. In practice, demographic urbanisation is measured and expressed, first of all, in the form of census-based statistics, the purpose of which is (at least in principle) is to distinguish the inhabitants of cities and towns from those living in rural areas [16]. The demographic dimension of urbanisation is related to the increase in the proportion of the urban population. Nevertheless, quantitative changes are also accompanied by qualitative ones. Changes are taking place in the gender and age structures of the population, the level of population growth, the balance of migration, the occupational structure of the population, the size of the family and its socio-professional status, and the size of the household, etc. [17] Accordingly, in recent studies, demographic urbanisation is examined in a much broader context, taking into account all the transformations occurring in demographic structures and processes, not only in quantitative but also in qualitative terms [4,18,19].

Spatial urbanisation is most often reflected in the rapid growth of cities, which has exerted pressure on land, local resources and especially on rural areas in their immediate vicinity [20,21]. The concentration of human activity in specific locations and regions

favors the development of large cities and various forms of urban settlement [22]. Urban growth and transformation are driven by the forces of attraction between specific locations. This process of spatial diffusion results from interactions between multiple factors and contributes to the development of new spatial patterns [23]. This is primarily due to the specific topography of the city surroundings, the existing road network, the land ownership structure and the pace of its development. These changes are catalyzed by the development of transport networks, as well as the advantages arising from the spatial agglomeration of resources, including human resources. Urbanisation is viewed as a process of organic growth, starting in the city center and leading to the spontaneous formation of 'mini' urban areas, especially throughout the transport network [24]. The majority of these processes are very dynamic and cause rapid changes in the structure and organization of land use [23]. The process of urbanisation, particularly spatial urbanisation, requires the application of continuously evolving research methods and techniques, which render it possible to grasp its essence and pace. The aim of this paper is to present the most interesting, according to the authors, methods used in research on the dynamics of the spatial urbanisation processes. Sources of information employed in the analyses of urbanisation processes are described and a number of examples of the use of spatial data in analyses of urbanisation processes are presented. The solutions described apply to multiple countries and continents and the issue is a common one. Studies addressing this subject usually test new methods for a universal understanding of the urbanisation phenomenon. The added value of this paper is in the fact that it describes ready-to-use solutions in a single publication, in the form of data sources and their use in analysing urbanisation processes. This should facilitate the selection of an appropriate method and its application in research on the dynamics of urbanisation processes for many researchers. The paper presents an overview of recent research and the most interesting, according to the authors, methods used in research on the dynamics of urbanisation processes. The main aim of the article was to produce a compendium that would guide the reader through the wide range of topics and provide inspiration for their own research. The first chapter defines the concept of urbanisation. Chapter 2 presents demographic changes and the spatial dimension of urbanisation from a global perspective. Chapter 3 describes recent trends in research on urbanisation processes. Sources of information used in the analyses of urbanisation processes are described in detail in Section 3.1. Section 3.2 presents examples of applying spatial data in analyses of urbanisation processes. Section 3.3 examines the possibilities of using spatial data in the study of urbanisation processes. Discussion, conclusions and further directions for research are presented in Chapter 4.

#### **2. Demographic Changes and the Spatial Dimension of Urbanisation from a Global Perspective**

Demographic processes are the driving force behind urbanisation. According to data published in World Urbanisation Prospects—The 2018 Revision [1]—approximately 55% of the world's population lived in towns and cities in 2013. It is projected that this value will reach 68% by 2050. This is due to changing lifestyles and a gradual reduction in agricultural employment in favor of the endogenous functions of cities [25]. According to the data published by the United Nations, two-thirds of the world's population will live in urban areas (Figure 1).

**Figure 1.** Global urban and rural populations: historical and projected. Source: [1].

In the second half of the 20th century, urbanisation rapidly accelerated. Dynamic growth of cities was observed in the more developed regions of the world, such as Europe and later North America. Yet, in recent decades, it has been Asia that has been urbanising at a tremendous rate—it now has more than half of the world's 40 megacities (with more than 10 million inhabitants). Africa is also gaining urbanisation momentum and now has three megacities in Cairo, Lagos and Kinshasa [26]. People migrate in search of a higher standard of living and the benefits of an urban lifestyle [27–29]. This process is particularly visible in Asia and Africa (90% urban population growth), where the highest rates of population growth are also found [30,31]. The migration of people to towns and cities was particularly high after 1950. From the level of 751 million, the urban population grew to 4.2 billion in 2018. It is predicted that by 2050 this number could rise by another 2.5 billion [32] (Figure 2).

**Figure 2.** Growth rates of urban agglomerations by size class. Source: https://population.un.org/ wup/Maps/ (accessed on 2 August 2021).

Today, the most urbanised areas include North America (82% of the population live in urban areas), Latin America and the Caribbean (81%), Europe (71%) and Oceania (68%). Although these percentages are much lower in Asia (50%) and Africa (43%) in view of the relatively lower numbers of cities and towns, these are the areas with the largest cities (over 10 million) [1]. It is expected that by 2030, the number of cities with over 10 million inhabitants will rise to 43, while in 1970, there were only three such locations. The scale of these changes is shown in Figure 3.

**Figure 3.** The projected growth of urban population and cities in 2030 compared to 1970. Source: https://population.un.org/wup/Maps/ (accessed on 2 August 2021).

Urban areas are the focal point of economic activity in most areas of the globe. Cities provide driving forces for development, bringing economies of scale, developing markets, creating jobs and encouraging new forms of business. As economies progress from basic activities—agriculture, fishing and mining—to industrial production and then services, the role of cities in the global economy expands with each transition. At present, about 55% of the world's population (4.2 billion people) live in cities, and by 2050, almost 7 out of 10 people worldwide will live in urban areas. According to the figures provided by the World Bank, more than 80% of global GDP is generated in cities. Major urban areas, particularly in developed countries, are economic giants [33]. The 600 largest urban centres, with a fifth of the world's population, generate 60 per cent of global GDP. The 380 largest cities in the top 600 in terms of GDP accounted for 50 per cent of global GDP [34].

Research studies conducted on the phenomenon of urbanisation are usually based on an analysis of statistical data concerning demographic and investment growth in administrative units. They are very significant as they clearly show the scale of the phenomenon. However, these studies often lack a spatial reference indicating the direction and the territorial extent of the changes.

The development of towns and cities is no longer controlled. The main reason behind this is the scale of economic and social change. The increased population mobility, the development of technology and the exchange of information have also resulted in city boundaries becoming indistinct [35]. The constantly growing urban population needs more space and the rate of growth results in a low level of space urbanisation processes [20,36]. This poses a serious threat to the spatial order and ultimately reduces the quality of this space and, consequently, the quality of people's lives [37]. Housing land, which accounts for over 70% of land use in most urban areas, determines the form and density of cities, provides employment and contributes to their growth. Housing policy and the provision of adequate infrastructure to residents for their safety and health is one of the most important challenges for these cities. In addition, it should be remembered that the shortage of

space causes price increases in the property markets, which leads to an increase in the number of residents living in extremely poor conditions or slums [22,38]. It is estimated that 881 million urban dwellers live in slums, a number that has increased globally by 28% in the last 24 years. Although the share of urban populations living in slums has declined over the past two decades, the number of slum dwellers continues to rise [24,39].

Forecasts show that the amount of urban land on Earth by 2100 may range from around 1.1 million to 3.6 million km<sup>2</sup> depending on the approach and the application of sustainable development principles—Figure 4.

**Figure 4.** Projected growth of urban areas between 2000 and 2100. Source: https://population.un. org/wup/Maps/ (accessed on 2 August 2021).

Transformations of land use triggered by new human demand affects many aspects of the environment at spatial and temporal scales, including freshwater quality and availability, extreme precipitation and flooding, biodiversity and habitat loss, and global warming [40]. Although cities around the world are the driving forces for economic value creation and income generation, fulfilling an essential role in many aspects of society, the spatial changes they induce and do not control can pose a huge threat to human beings.

Uncontrolled urban growth is clearly visible in the surroundings of metropolitan cities and often results from insufficient space that can be allocated for investment purposes within the city to satisfy the needs of an economically and demographically expanding metropolis. The spatial dimension of urbanisation applies to both urban and rural areas, which is reflected in the amount of space used in a characteristic urban manner. Rural areas, in particular, are experiencing a constant transformation, which is caused by the interest in these areas of urban dwellers. These processes are usually very dynamic and produce rapid changes, based mainly on the principle of reorganisation of the structure of space, and manifest themselves as advanced suburbanisation and exurbanisation [35]. In the literature, this phenomenon has become known as 'urban sprawl', denoting the process of spreading and the enlargement of big cities [41,42]. This phenomenon is seen as a process resulting from socio-economic changes due to spontaneous and disorderly urban expansion [43]. For this reason, it is extremely important to monitor the direction, the level and the pace of spatial urbanisation. Urban sprawl is often measured by the population density gradient (as the percentage decrease in density over increasing distance from the city centre) [44] or by the spatial size of the urban area [45]. The analysis of these values shows a disproportionate increase in urban space to the increase in urban population [46].

Suburbanisation is of particular importance in a period of rapid urban growth. Areas under direct urbanisation pressure are described in the literature as the urban–rural transition zone [47,48], the urban–rural continuum [49–53] or the suburban zone [54–57]. Studies of these areas focus not only on the progressive integration of new land into the urban sphere of influence, but also on changing land use, the extent of infrastructure, access to services and markets, and exposure to urban production processes and environmental

pollution [58]. Changes in land use in areas are referred to as a rural–urban fringe [59–64], in so-called "green belts" surrounding the city [65–67] and areas referred to as "urban villages" [68] are also explored. Areas under urbanisation pressure are also investigated in terms of investment attractiveness [69,70] or real estate market analyses [71–75].

#### **3. Methodology of Proceeding and the Determination of Recent Trends in Research on Urbanisation Processes**

In order to describe the investigated area of knowledge, based on an extensive literature review, a number of methods and techniques used to study the dynamics of urbanisation processes were examined. The variety of attributes and the large range of data make the process of analysing spaces under urbanisation pressure both complex and time-consuming. The key to the choice of the spatial data presented and the methods applied in the research on the dynamics of urbanisation processes was their availability, the difficulty of classifying certain forms of development, the time of implementation, costs and their spatial and temporal continuity. The main focus of the paper is on the use of photogrammetric and remote sensing sources to study urbanisation changes. Among these, sources were separated according to the type of information and their spatial scope. The examined sources demonstrate land cover changes, which can be divided into four classes: land-use change analyses, environmental change analyses and demographic and economic change analyses. The spatial extent of the data gathered is highly diverse. Its scope can be classified as local, regional, national or global. Taking into account these two aspects, the most interesting examples of applying particular databases created with photogrammetric and remote sensing sources for the purpose of analysing the dynamics of urbanisation processes were identified by a literature review. The presented research results are based on data from 1990–2020. The classification of spatial data sources and the selection procedure of application cases are shown in Figure 5.

**Figure 5.** Classification of spatial data sources.

#### *3.1. Sources of Information Used in Analyses of Urbanisation Processes*

Urbanisation produces profound changes in space and gradually modifies the structure of land use. There is a great need to monitor urban space, as well as the changes that are taking place in areas subject to the direct pressure of urbanisation. The development of science and technology provides many new tools to observe urbanisation processes and to analyse and draw conclusions regarding this phenomenon. Sources of information data used in analyses of urbanisation processes can be generally divided into statistical, cadastral, photogrammetric and remote sensing.

Statistical data used in analyses of urbanisation processes take into account changes occurring over time in land-use forms, settlement unit boundaries, changes in the territorial division, in the number of towns and cities, as well as in the area, population and structure. Statistical data collected and processed, e.g., by the European Statistical Office (Eurostat) and governmental statistical offices (Statistics Poland), also refer to internal and external population migration processes.

Cadastral data used in an analyses of urbanisation processes are gathered in public registers and cartographic studies. The data in the registers include, among others, information on persons, their rights, legal transactions, obligations, goods, etc. For example, in Poland there are about 280 public registers serving different purposes, e.g., the registration of the actual state of real estate is assigned to the land and building register, while the registration of the legal status of real estate belongs to the scope of the land and mortgage register [76]. Descriptive and spatial attributes of real estate used in an analyses of urbanisation processes include the number of parcels, numerical description of boundaries, surface areas, land use contours and soil classes, designation of the land and mortgage register or sets of documents defining the legal status of the real estate, street names and ordinal numbers of buildings, names of physiographic objects, GESUT (geodetic records of public utilities) data, providing information on ground, aboveground and underground technical infrastructure. Data from cartographic studies also contain information on the condition, distribution and interrelationships of various types of terrain objects and phenomena, their qualitative and quantitative characteristics, as well as their names and descriptions. The analysis of urbanisation processes with the use of digital or analogue cartographic works, such as a topographic map, takes into account surface elements of the landscape, such as residential estates, industrial and agricultural facilities, water and related facilities, vegetation, crops, land, boundaries, landforms, etc. In general, cadastral data can be divided into qualitative (describing the type, equipment, buildings, utilities), quantitative (area, volume) and situational information (location, access).

Photogrammetric and remote sensing data seem to be particularly important in the analysis of urbanisation processes. Photogrammetry (terrestrial and aerial) is used to develop and update various types of maps, orthophotomaps, digital elevation models and spatial information systems. Remote sensing data (data on phenomena and objects acquired without direct contact with them), which are produced at various altitudes and by various sensors, can be divided into active (radars-SAR, scanners-LiDAR) and passive data (cameras or thermal scanners-placed on the board of satellites, aircraft and drones). The sensors provide images in various ranges of electromagnetic radiation. The resulting databases, complemented by spatial databases, which are already common today, provide enormous research material and make it possible to study the phenomenon of urbanisation with greater intensity and in a more up-to-date context.

Photogrammetric and remote sensing (RS) data and GIS tools provide accurate information on land use and land cover changes. Currently, the main sources of data are the Copernicus (program supervised by ESA) and Landsat (program supervised by NASA) satellites. The data acquired and shared from the Copernicus (in particular Sentinel-2 and Sentinel-1) and Landsat satellites provide a high level of detail. The use of a variety of sensors, including optical and radar, facilitate monitoring land cover and support crisis management. Additionally, the photogrammetry and remote-sensing products obtained through the Copernicus and Landsat programs, generating high-resolution images of land

cover [77], make it possible to monitor urbanisation processes [78–80]. Monitoring land cover with high-resolution imagery, especially in urban areas, is a key task and has applications in many fields, such as land development planning, urban planning and architecture, ecology and environmental protection, etc. These data are also of fundamental importance for understanding spatial urbanisation processes. With the availability of urbanisation data, photogrammetry and remote sensing techniques are becoming increasingly popular for monitoring changes in land use processes. The Copernicus program services, coordinated by the European Environment Agency, provide information on land cover and land use on a European, national and local scale. The pan-European component includes, among others, the uniform land cover databases (CLC1990, CLC2000, CLC2006, CLC2012 and CLC2018) and High Resolution Layers (HRL).

The CORINE Land Cover database includes all land cover forms occurring on the European continent, leaving no unclassified areas (clc.gios.gov.pl, access date: 1 February 2018). The database contains information on both land cover and land use. In the CLC, land cover forms are hierarchically divided into three levels. The database distinguishes five basic forms of land cover: anthropogenic land (built-up areas, used for housing, services, industry or mines, and municipal green spaces), agricultural land (arable land, permanent crops, meadows and pastures, and wooded and shrubby land used for agriculture), forests and ecosystems (land covered with or partially devoid of forest vegetation), wetlands (inland marshes, peat bogs, salt marshes and mudflats) and water areas (inland waters and sea waters). These forms are further subdivided into levels two and three, which specify the form of land use within the group [81]. The CLC database is a tool for conducting complex spatial analyses based on diverse land-use types. One of the greatest advantages of the CLC dataset is that it is regularly updated, which makes this resource particularly useful for analysing the rate of changes in land use and developing forecasts. HRLs contain detailed information on characteristic land cover forms: impervious areas, wooded areas, areas with grass cover, water bodies and wetlands. These layers are complementary to the Corine Land Cover databases.

The local component also offers Urban Atlas, Riparian Zones and Natura 2000. The Urban Atlas data represents the functional zones of urban areas and contains detailed data on land cover (land use) compiled for the most populous European cities (for most cities with more than 50,000 inhabitants). The classification recognises 17 urban classes (including five classes of mixed-density development) and ten other classes associated with other land cover forms. The Urban Atlas project currently comprises nine products, including land cover data from 2006, 2012, 2018 and Urban Atlas Change 2006–2012 and 2012–2018. Riparian Zones data concern land cover and land use in areas along rivers, i.e., riparian areas. The primary objective is the need to monitor biodiversity at the European level, including as part of the improvement of 'green' and 'blue' infrastructure in the European Union. Natura 2000 data relate to grassland-rich areas and their assessment for conservation efficiency. Coastal Zones data are used to monitor the dynamics of land-use change in coastal zones.

The above-mentioned types of sources of information on urbanisation processes are merged to create Spatial Information Systems (SIP). The data necessary for analyses of urbanisation processes can be found in various databases of this system, such as Euro-Geographics. EuroGeographics facilitates access to official, comparable and verifiable geospatial data from European national authorities responsible for maps, cadaster and land and mortgage registers. EuroGlobalMap is a geodatabase at a 1:1,000,000 accuracy level covering 45 European countries and territories. The thematic scope of the data includes administrative boundaries, transport and hydrographic networks, landforms, settlements and anthropogenic objects. EuroGlobalMap provides an ideal background for a wide range of activities, from network planning, monitoring and analysis to the presentation of environmental policies.

Access to key data is also provided by all types of national spatial information systems. Spatial information refers here to information on the location (coordinates in an

assumed reference system), geometric properties, and spatial relationships of objects that are identified in relation to the Earth and that can be used in an analyses of urbansation processes. In Poland, for example, the National Geographic Information System (KSIG) consists of standardised reference databases containing information on objects located on and below the Earth's surface, together with their location, situated in the territory of Poland, as well as procedures and techniques for the systematic collection, updating, processing and disseminating of the data. The most important components of KSIG include vector map level 2 (VMAP2), topographic database (TBD), topographic and base maps, general geographic database (BDO), photogrammetric database, orthophotomaps and the land and building register (EGIB). The most important feature distinguishing the spatial information systems in operation, which is the consequence of the presence of spatial information contained in them, is the possibility of its cartographic presentation and of an analysis allowing answers to be obtained regarding the real world modelled by the system [82].

#### *3.2. Examples of Using Spatial Data in Analyses of Urbanisation Processes*

The examples presented below reflect the latest and, in the authors' opinion, the most interesting trends in the research on the dynamics of spatial urbanisation processes.

#### 3.2.1. The Analysis of Urbanisation Processes Using Aerial and Satellite Images

Aerial and satellite imagery is very often used in spatial analyses. The result of processing these images is an orthophotomap, which, unlike a traditional map, presents the state of land cover and land use in the most realistic way. The use of publicly available orthophotomaps enables the identification and location of the urban investment boundary, the degree and rate of urbanisation of the area (Figure 6).

**Figure 6.** An orthophotomap of the southern part of Olsztyn: (**a**)—1995, (**b**)—2005, (**c**)—2009, (**d**)—2017. Source: https://msipmo.olsztyn.eu/imap/ (accessed on 20 July 2021).

The expert method of visual interpretation of the orthophotomap (photointerpretation) and additional data—e.g., from the land and building register and from the field visits, make it possible to identify current and historical land-use forms (Figure 7).

**Figure 7.** Inventory of the existing development—an example, 2017. Source: [20].

At present, orthophotomaps are produced with a field resolution of 10 cm, which allows the precise identification of objects. Orthophotomaps are successfully used as separate layers in geographic systems and form the background for spatial planning and the creation of thematic studies.

#### 3.2.2. The Analysis of Urbanisation Processes Using Corine Land Cover Data (CLC) Data

The European Earth observation program known as Copernicus Land Monitoring provides two datasets on land cover characteristics in Corine Land Cover data (CLC) and High Resolution Layers (HRL). The Corine Land Cover initiated in 1985, updated in 2018, provides information for the whole area of the European Union—39 countries. Mainly, the identification of land cover is mapped by a visual interpretation of high-resolution satellite images. The CLC dataset is divided into 44 classes, which describe five categories: artificial surfaces, agricultural areas, forest and seminatural areas, open spaces with little or no vegetation and wetlands and water bodies [23,48,83,84].

A land cover model based on Corine Land Cover data (CLC) is presented in a study conducted by Biłozor A. et al. (Identification and Location of a Transitional Zone between an Urban and a Rural Area Using Fuzzy Set Theory, CLC, and HRL Data). The result of the aggregation of individual CLC data classes defining the range of land uses for the selected area is shown in Figure 8.

The CORINE Land Cover (CLC) databases compiled for the period 1990–2000–2006– 2012–2018 provide another reliable source of information about ongoing urbanisation processes. One of the uncontrolled urbanisation phenomena is suburbanisation connected with the emergence of urbanised areas far beyond the city limits in the form of often chaotically located buildings. This is common practice but is undesirable due to the quality of these areas, which tend to be poorly equipped with technical infrastructure and significantly increase the cost of commuting to work and of basic services. An example of using the CLC database, Geographical Information System (GIS) tools and the overurbanisation (OU) indicator described in the publication: The Use of the CORINE Land Cover (CLC) Database for Analyzing Urban Sprawl [84] is presented in Figure 9.

**Figure 8.** Land cover model in the area under investigation based on the CLC data. Source: [48].

**Figure 9.** Aggregated urban areas. Source: [84].

3.2.3. The Analysis of Urbanisation Processes Using High Resolution Layers Data (HRL)

Another type of data-High Resolution Layers (HRL) provides more detailed information about land cover than CLC. This dataset was first produced in 2012 from satellite imagery through a combination of automatic processing and interactive rule-based classification. Currently, the main source of data are satellites created in the Sentinel project, in particular, Sentinel-2 and Sentinel-1, which allows the use of different sensors, including optical and radar. In this dataset, which is presented in spatial resolution of 20 × 20 m for 39 countries in the EU, five themes can be identified which correspond with the main categories in CLC. The level of sealed soil (imperviousness degree 1–100%) which is produced

using a semi-automated classification based on calibrated NDVI, captures the different types of specific land cover. Those five products capture the spatial distribution of Imperviousness, Forest, Natural Grassland, Wetlands, Permanent Water-bodies, Wetness and Water and Small Woody Features [85–87]. The land cover model based on HRL Imperviousness data is presented in Biłozor A., Czyza Sz., Bajerowski T.-Identification and Location of a ˙ Transitional Zone between an Urban and a Rural Area Using Fuzzy Set Theory, CLC, and HRL Data [48] (Figure 10). The conducted analyses included raster reclassification followed by the polygonization of urbanised areas. The measures taken enabled the indication of the boundary of the urban area, which, according to the adopted assumptions, determined areas with imperviousness at a level of 30%.

**Figure 10.** Land cover model in the area under investigation based on HRL Imperviousness data Source: [48].

HRL databases are also a reliable source of information about ongoing urbanisation processes. In their research, Liua X. et al. (High-resolution multi-temporal mapping of global urban land using Landsat images based on the Google Earth Engine Platform) used the Normalized Urban Areas Composite Index (NUACI) method and utilised the Google Earth Engine to facilitate the global urban land classifications from an extensive number of Landsat images [87]. High-resolution multi-period mapping (with a 5-year interval) of global urban areas using Landsat imagery based on the Google Earth Engine platform is shown in Figure 11.

#### 3.2.4. The Analysis of Urbanisation Processes Using Urban Atlas Data

The Urban Atlas contains pan-European comparable data on land cover and land use comprising a range of functional urban areas. The presented example describes changes to the urban landscape identified and assessed using the Urban Atlas data. The aim of the research conducted by Pazúr R. et al. (Changes of urbanised landscape identified and assessed by the Urban Atlas data: Case study of Prague and Bratislava [88]) was to document, examine and compare changes in land use/cover (LUCC) of LUZ (Large Urban Zones) in Bratislava and Prague in 2006–2012 using the UA data, and to demonstrate how these changes are recorded in official statistics—Figure 12.

**Figure 11.** Urban land expansion from 1990 to 2010 in the representative cities. Source: [87].

**Figure 12.** (**a**) LUC in (A) FUA Bratislava and B) FUA Prague in the years 2006 and 2012 according to the UA data; (**b**) LUC gains in (A) FUA Bratislava and (B) FUA Prague in the period 2006–2012 according to the UA data. Source: [88].

#### 3.2.5. Monitoring Changes in Land Use with Landsat Time Series

The presented example shows the mapping of urban development patterns in Ouagadougou, Burkina Faso, using machine learning regression modelling with Landsat two-season time series [89]—Figure 13. The study was designed to quantify land cover for the Ouagadougou metropolitan area from 2002 to 2013 using Landsat-TM/ETM + /OLI time series.

#### 3.2.6. The Use of Nighttime Light (NTL) Data in Analyses of Urbanisation Processes

Nighttime Light (NTL) data are increasingly often used in an analyses of urbanisation processes because of their strong relationship with human activities. Systems such as the Visible Infrared Imaging Radiometer Suite (VIIRS) Day/Night Band (DNB) onboard the Joint Polar Satellite System (JPSS) are designed with very high sensitivity to detect even the faintest light in the visible range. This makes them ideal for observing the lighting on the ground. An example of the use of these data is presented in The Annual Cycling of Nighttime Lights in India [90]—Figure 14. By classifying the ACF (autocorrelation function) profiles for each pixel location, interesting spatial patterns were revealed in different regions.

**Figure 13.** Fraction development of urban surfaces, seasonal vegetation and soil from 2002 to 2013. Source: [89].

**Figure 14.** (**a**). Land use/land cover map for India. (**b**). Binary background mask from the 2019 annual VIIRS Nighttime Light (VNL) product. Source: [90].

Up-to-date and accurate information on the dynamics of urban expansion is essential to reveal the relationship between this phenomenon and the ecosystem in order to optimise land-use patterns and to promote efficient urban development. A publication entitled "Extracting the dynamics of urban expansion in China using DMSP-OLS nighttime light data from 1992 to 2008" [91] describes a method for systematically correcting multi-year stable night light (NSL) data from multiple satellites and rapidly extracting urban expansion dynamics based on the corrected data—Figure 15.

**Figure 15.** Accuracy assessment of selected urban areas in China in 2008 extracted using NSL data. Source: [91].

The dynamics of city expansion in China from 1992 to 2008 were extrapolated with an overall average accuracy of 82.74% and a mean Kappa value of 0.40—Figure 16.

**Figure 16.** The dynamics of urban expansion in China from 1992 to 2008. Source: [91].

3.2.7. The Use of Greenhouse Gas Emission Data in Analyses of Urbanisation Processes

The research conducted by Christopher Jones and Daniel M. Kammen in the USA revealed that the spatial distribution of household carbon footprints identifies a process of suburbanisation by analysing the effects of greenhouse gas emissions linked to urban density [92]. The econometric models of demand for energy, transportation, food, goods and services were developed using national household surveys and were used to determine the average household carbon footprint (HCF) for U.S. zip codes, cities, counties and metropolitan areas. A lower HCF was observed in city centres and a higher carbon footprint in the peripheries (~50 tCO2e), ranging from ~25 to > 80 tCO2e in the 50 largest metropolitan areas. Population density demonstrates a weak, though positive, correlation with HCF until a density threshold is reached, after which the range, mean and standard

deviation of HCF decrease. Population density contributes to the relatively low HCF in the urban centres of large metropolitan areas. In contrast, more extensive suburbanisation in these regions contributes to an overall net increase in HCF—Figure 17.

**Figure 17.** HCF from (**A**) electricity, (**B**) natural gas, (**C**) fuel oil and other fuels, (**D**) housing = A + B + C + water, waste, and home construction, (**E**) transportation, (**F**) goods, (**H**) food, (**I**) services, and (**G**) total = D + E + F + H + I. Transportation includes motor vehicle fuel, lifecycle emissions from fuel, motor vehicle manufacturing, air travel direct and indirect emissions, and public transit. Scales below each map show gradients of 30 colours, with labels for the upper value of lowest of quantile, median value and lowest value of highest quantile, in metric tons CO2e per household, for zip code tabulation areas (ZCTAs). East Coast metropolitan statistical areas (**J**), with a larger map of New York metropolitan area (**K**, outer line) and New York City (**K**, inner line), highlight the consistent pattern of relatively low GHG urban core cities and high GHG suburbs. Source: [92].

#### 3.2.8. The Use of Data on Plant Vegetation Changes in Analyses of Urbanisation Processes

Urbanisation destroys and divides large amounts of natural habitats, with serious consequences for ecosystems, which is particularly evident in developing countries. The term 'urban landscape' refers to a region where specific landscape elements within an urban area prevail, including buildings, roads, infrastructure and green spaces. A number of studies have been carried out to monitor the dynamics of urban landscape changes and the ecological impact of urbanisation. Sufficient evidence has been found to demonstrate that changes in the spatial arrangement of urban landscapes (e.g., distribution, composition and configuration) lead to mass plant die-offs [93], a decrease in biodiversity [94], a reduction in atmospheric humidity [95] and a change in ground surface temperature [96,97], and an

increase in fine particulate pollution [98]. Therefore, the change in the urban landscape pattern is one of the most striking aspects of urbanisation, with a negative impact on ecosystems. The presented example describes the response of vegetation to a change in the urban landscape spatial pattern in the Yangtze River Delta, China [99]. Figure 18 depicts the spatial patterns of vegetation cover in 2004, 2008, and 2013 in the Yangtze River Delta Urban Agglomeration (YRDUA). The remote sensing datasets used in this study include the Defense Meteorological Satellite Program's Operational Linescan System (DMSP/OLS) and the normalised difference vegetation index (NDVI) obtained from a Moderate Resolution Imaging Spectroradiometer (MODIS) dataset. The vegetation level with a *fc* (fractional vegetation cover) value higher than 0.75 extended over a large part of the study area, which was mostly dominated by trees, shrubs or green plants. Scattered vegetation in the northern and central parts of the YRDUA experienced the most notable reduction. It should be noted that since these two regions mainly consist of arable land and building land, the decline in *fc* can be attributed to land use transformation caused by intensive human activities, and vice versa, as the *fc* values in the southern part of the YRDUA, i.e., most of Zhejiang Province, have basically remained unchanged. Until 2013, the *fc* value for the entire area under investigation presented a clear spatial disproportion between the north and the south.

**Figure 18.** Spatial patterns of vegetation cover for (**a**) 2004, (**b**) 2008, and (**c**) 2013. Source: [99].

The spatial pattern of the vegetation change trend is shown in Figure 19.

**Figure 19.** Trends of vegetation change in (**a**) YRDUA and (**b**) urban sprawl area in 2004–2013. Source: [99].

3.2.9. The Use of Synthetic Aperture Radar (SAR) in the Analysis of Urbanisation Processes

Synthetic Aperture Radar (SAR) for high-resolution images of stationary objects is used to create images of the land surface, of the Earth and other planets using remote sensing techniques. Remote sensing satellites, especially the German TerraSAR-X radar system, independent of weather and time of day, and which takes up to three days to revisit a particular area, have been successfully providing short-term data on urbanisation processes for many years. The presented example investigates the potential and limitations of TerraSAR-X in the context of automated, object-based detection of human settlements. The approach presented by Thiel M. et al. (Object-oriented detection of urban areas from TERRASAR-X DATA) [100] for urban detection using TerraSAR-X data, achieves an overall accuracy of around 95% and 89%—an indication of the high potential of TerraSAR-X data to identify developed areas—Figure 20.

**Figure 20.** Filtered TerraSAR-X intensity layer (**a**) and calculated speckle divergence layer (**b**) of a subset of the region of Munich. The built-up areas are characterised by high values of speckle divergence. Source: [100].

#### *3.3. Possibilities of Using Spatial Data in the Research on Urbanisation Processes*

The development of a relatively simple measure of urban sprawl based on the observation of changes in actual land cover using GIS and a tool supporting rational land management provides a substantive basis for city planning. An example is the research described in a study entitled "An analysis of urbanisation dynamics with the use of the fuzzy set theory—A case study of the city of Olsztyn", enabling the application of fuzzy set theory as a tool supporting rational space management and provides a substantive basis for urban spatial planning [20]. The established degrees of membership determine which forms of space use have a so-called "more city-like" character than the others, i.e., have more characteristics of urban space. The results according to the assumptions of the fuzzy set theory are determined in the interval [0, 1]. The determined degrees of membership in urban-type uses and the data from the inventory of the existing state of development (interpretation of satellite images) in 2005, 2010 and 2018 enable the development of a fuzzy city model—Figure 21.

Boundaries of urbanised areas determined by the developed method based on the fuzzy set theory for 2005, 2010 and 2018 make it possible to determine the degree of urbanisation of the area in the interval [0, 1] and the dynamics of changes in urbanisation processes in the years of 2005–2010–2018 (Figure 22).

**Figure 21.** Urban use in the study area in the degree of membership interval of [0.00–1.00]: (**a**)—2005; (**b**)—2010; (**c**)—2018. Source: [20].

**Figure 22.** Changes to land use in the years: (**a**): 2005–2010; (**b**): 2010–2018; (**c**): 2005–2018. Source: [20].

The spatial development of cities and the related population growth has become a subject of broadly understood spatial management research. Planners and geographers are replacing previous models of cities that described the process of their development with new models that express how uncoordinated local decisions affect their global growth [101]. Urban compactness studies can identify cities which are expanding to make optimal use of suburban areas and those where suburban sprawl is chaotic. An example of research into the compactness of urban systems is the establishment of a compactness index based on the spatial length of the boundaries of the area covered with a particular form of development [84]. Studies of this type can apply GIS-type data; in this case, the Corine Land Cover Data CLC databases. Such data, for example, was used to examine the compactness of the urbanised areas of 49 district towns in Poland. For the 49 urban areas, the zone of the neighbouring municipalities was separated as the area most exposed to suburbanisation, and subsequently, using the CLC databases according to the developed research methodology, the areas related to urban use were separated for the three moments of time 2006, 2012 and 2018—Figure 23.

**Figure 23.** The areas related to urban use separated for the three moments of time 2006, 2012 and 2018. Source: own elaboration.

Based on the data concerning the area of these zones, the area compactness index AC was calculated, which is equal to the ratio of the circumference of a circle with the surface area equal to that of the urbanised areas to the length of the boundaries of these zones. The index ranges from 0 to 1. The values closer to zero indicate lower compactness. The value of 1 is theoretical and would represent the area of the city with an urbanised area forming a dense circular surface.

This indicator increases for most cities in 2018 as compared to previous years, indicating that areas that were heavily dispersed between 2006 and 2012 are becoming more compact—Figure 24. Unfortunately, for all towns, the index for the years 2006, 2012 and 2018 included in the analysis is very low (apart from 4 examples, it does not exceed the value of 0.4, and for more than a half, it is below 0.2), which indicates a strong dispersion of development in the vicinity of the towns under analysis. The examples provided in the discussion illustrate how spatial data can be used to assess the level and pace of urbanisation. As spatial databases become more accurate, complete and accessible, with images taken and made available at a better resolution, more refined and novel methods for mapping and analysing urbanisation processes will emerge.

**Figure 24.** The AC index in the analysed cities in Poland. Source: own elaboration.

#### **4. Discussion and Conclusions**

The dynamics of urbanisation processes have been the subject of multiple studies differing in both scale and aspect. Some interesting publications on this subject include works

which are well known to researchers in this field, such as [102], but also some other very interesting items [103,104]. These include some studies that use land cover information, such as [105,106]. In the majority of these papers, the authors identify specific methods for environmentally, economically or demographically oriented research. Contrary to other manuscripts of this kind, this paper shows the wide possibilities of using spatial data to study urbanisation processes at different scales and with different underlying meaning.

The analytical methods and databases discussed above differ in their availability, ease of classification of selected land uses, labour intensity, cost, spatial and temporal continuity. Most frequently, they also require a specialised knowledge and dedicated algorithms to process the collected data. Automatic classification can result in errors in identifying and separating different land-use types when a large number of similar and overlapping land cover types are processed, such as single-family homes and multi-family housing. In view of the variety of colours, structures and textures in urban areas, spatial conditions in the vicinity of the identified areas should also be taken into account. The interpretation of aerial and satellite photographs alone is time-consuming and requires a team of several researchers. However, from a longer perspective, such analyses are less time- and costintensive and can be carried out on a large number of cities to generate sufficient data to draw conclusions concerning the degree and nature of urbanisation.

Urbanisation is a global phenomenon, irreversible and inherent within human development. The inevitability of these processes stimulated by developing cities requires systematic research into the characteristics and parameters that form them and the various processes of transformation taking place. Urban development significantly affects areas located at the interface between urban and rural uses, resulting in the permeation and overlapping of different land-use forms in the fringe zone. Spatial development of cities manifested by an increased demand for new land exerts a significant impact on the surrounding grassland and agricultural areas. The demand for new non-urbanised building land, much cheaper than property in cities, is triggering a response from municipal governments, environmentalists, farmers and the non-agricultural rural population. Identifying the specific nature of urbanisation processes occurring in and around cities can minimise the uncontrolled and unplanned creation of zones of chaotic development, degrading natural resources and could limit uncontrolled and unplanned city growth. This is also of fundamental importance in the context of sustainable urban development. New methods of urban management, environmental protection, uncontrolled suburbanisation, acquisition of new urban land (agricultural as well as post-industrial) transport, social participation, etc., are new challenges for sustainable urban areas and practices. Planning for sustainable urban development requires a knowledge of the individual elements of the city system and its surroundings and the relationships between them, the adoption of certain assumptions and goals leading to sustainability, and the adaptation of these assumptions to local conditions. This is particularly important from a spatial perspective [107–114].

Rapid urban growth and the resulting challenges require precise techniques for identifying and locating urbanised areas and for their mapping in order to represent complex land cover characteristics in adequate detail. Geospatial data provide a rich source of knowledge and a reference for the development of new tools for identifying and monitoring urbanisation processes. This applies in particular to areas subject to urbanisation pressures which cannot be clearly classified as urban or rural. These discrepancies are reflected in the existing land use and land cover types. For this reason, the degree and rate of urbanisation can be most effectively analysed by monitoring land cover changes. Virtually unlimited access to land cover data provided by current and historical orthophotomaps and databases, as well as developing GIS techniques, enable the rapid processing of spatial information. Research on spatial urbanisation consumes much less time and concerns a large number of cities, which may provide new sources for drawing conclusions about the scale and nature of this phenomenon.

The possibilities and constraints related to decision-making regarding the spatial development of the city are primarily attributable to the multidimensional nature of the

spatial development space and its probabilistic and fuzzy nature. The variety of spatial features (attributes) and the wide range of data result in the process of investigating and planning a suburban area, which is complex and lengthy and, consequently, subject to high risk. Studies of urbanisation processes must take into account the nature of the explored component, characteristics, as well as the frequency, pace and magnitude of these changes. The analyses presented in this paper provide an attempt to further define current trends, methods, techniques and databases used in the analysis of the ongoing urbanisation processes.

**Author Contributions:** Conceptualisation, A.B. and I.C.; methodology, A.B. and I.C.; software, A.B. and I.C.; validation, A.B. and I.C.; formal analysis A.B. and I.C.; investigation A.B. and I.C.; resources, A.B. and I.C.; data curation, A.B. and I.C.; writing—original draft preparation, A.B.; writing—review and editing, I.C.; visualisation, A.B. and I.C.; supervision, A.B. and I.C.; project administration, A.B. and I.C.; funding acquisition, A.B. and I.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **A Geo-Spatial Analysis for Characterising Urban Sprawl Patterns in the Batticaloa Municipal Council, Sri Lanka**

**Mathanraj Seevarethnam 1,\* , Noradila Rusli <sup>2</sup> , Gabriel Hoh Teck Ling <sup>1</sup> and Ismail Said <sup>3</sup>**


**Abstract:** Urban sprawl related to rapid urbanisation in developed and developing nations affects sustainable land use. In Sri Lanka, urban areas have mostly expanded in a rather spontaneous, unplanned manner (based on the current settlers' subjective movement) rather than conforming to the local government's development plan. This growth inevitably leads to uncontrolled urban sprawl in many Sri Lankan cities, including Batticaloa. So far, Sri Lanka's planners or researchers have not yet tackled the sprawling developments in this city. Understanding the different forms and patterns of urban sprawl is the key to address sprawling growth. This study aims to identify the characteristics of urban sprawl in the Batticaloa municipal council using Geographic Information System (GIS) and remote sensing technology. Landsat satellite images for the years 2000, 2010, and 2020 as well as 2002, 2011, and 2019 population data were used and analysed using ArcGIS' maximum likelihood classification tool and the density function, respectively, to delineate the characteristics of urban sprawl. The results revealed that low-density development, leapfrog development, commercial ribbon development, and scattered growth are the influencing characteristics of urban sprawl in the Batticaloa municipality. These characteristics were found mainly in the urban edge of the city and have led to urban sprawl. The finding provides knowledge into recognising the characteristics of urban sprawl with empirical evidence. It affords a clear direction for future studies of urban sprawl in rapidly growing cities that are numerous in Sri Lanka, and the identified characteristics of urban sprawl can be useful in minimising future sprawl. This result can be a tool for future urban planning and management in the Batticaloa municipality.

**Keywords:** urban sprawl; land use; urbanisation; leapfrog development; scattered development

#### **1. Introduction**

Urbanisation is a reflection of the human activities affecting the land that has been threatened by the enormous pressure from population growth [1]. Rapid urban growth is generally related to and driven by the concentration of population in an area [2]. According to the United Nations' world population prospects in 2019, there will be an increase in the next 30 years of two billion people, from the current world population of 7.7 billion to 9.7 billion in 2050. Further, this increase will grow to almost 400 cities in the early 21st century, which includes around 70% in developing countries [3], including Sri Lanka. Currently, the urban population of Sri Lanka is almost 25% of the total population, which is expected to increase by 65% by 2030, which will cause cities to grow physically and numerically, creating urban sprawl issues in the future [4].

Urban sprawl has attracted much attention among policymakers and researchers in developed [5,6] and developing [4,7–9] countries worldwide. Most arguments for

**Citation:** Seevarethnam, M.; Rusli, N.; Ling, G.H.T.; Said, I. A Geo-Spatial Analysis for Characterising Urban Sprawl Patterns in the Batticaloa Municipal Council, Sri Lanka. *Land* **2021**, *10*, 636. https://doi.org/10.3390/ land10060636

Academic Editors: Luca Salvati, Iwona Cie´slak and Andrzej Biłozor

Received: 15 May 2021 Accepted: 9 June 2021 Published: 15 June 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

urban sprawl are not based on strong empirical evidence but rather on speculation and assumptions [10]. Many researchers explain this concept with the urban environment of the research area. So far, there is no consensus on the definition of urban sprawl. Further, urban sprawl is a socioeconomic phenomenon that has gradually become a critical issue in many urban areas [11], including Sri Lankan cities. Built-up area, which is often considered a parameter for measuring urban sprawl [2], especially settlement density or size, can only afford an overall measure of the urban form [12]. Increasing urban sprawl results in household preferences, the locational choice for commercial investment, often being agricultural, and building construction in the cities [13,14]. Vacant lands, which are primarily transformed into housing on a daily basis, are increasing sprawling growth and property costs in the urban periphery and surrounding areas.

As a developing country in the world, Sri Lanka has been studied to define urban sprawl [4] and its impact [15] particularly in Colombo city and the spatiotemporal patterns of urban sprawl in Kandy city [8]. Besides these, the Batticaloa municipality area is highly accompanied by residential and commercial development. According to the Sri Lanka National Physical Planning Policy and Plan 2010–2030, Batticaloa is a rapidly developing city in Sri Lanka that is expected to become a metro city by 2030. Considering the existing situation in the Batticaloa municipality, many ongoing urban development projects have been carried out since the end of the Civil War in 2009. However, these developments were not well designed by planning experts. It led to less effective growth and changes in the urban area. Some studies have examined the urban land use changes in this area [16,17], and these studies provide an insight into why this area should be necessary for studying urban sprawl patterns. The built-up development has been growing rapidly in this area in recent decades [16], creating an irregular pattern or sprawling growth established with empirical evidence.

Furthermore, the characteristics of urban sprawl have been studied in various cities around the world (see Table 1), such as India, Malaysia, China, Romania, and the United States. However, these studies have not identified all the characteristics of urban sprawl in a single city. Several characteristics were only identified in a particular city in a developed country or a developing country. Moreover, since Sri Lanka is a developing country, it is difficult to identify influencing characteristics of urban sprawl based on the experience of previous literature from developing countries. Furthermore, cities in Sri Lanka never studied the characteristics of urban sprawl before, which also makes it difficult to understand the pattern of urban sprawl in the Batticaloa city. More precisely, there is still a lack of knowledge, especially about the characteristics (forms and patterns) of urban sprawl via the analysis of the different parts of the city, which is essential to tackle the sprawl effectively. The built-up patterns are the key parameter to identify the different characteristics such as low-density development, leapfrog development, commercial ribbon development and scattered development to establish the urban sprawl development. So far, the planners or academics have not yet addressed the sprawling development in this city in Sri Lanka. If this growth continues in this city, it will affect its sustainability when it becomes a metro city in 2030. In the end, this study can answer which characteristics influence the Batticaloa municipality through geospatial analysis.

Thus, this study can involve finding the different characteristics of urban sprawl in the Batticaloa municipality through spatial patterns. This finding can minimise the sprawling growth in the future and develop a sustainable city in Sri Lanka. The influencing characteristics in this city can be identified from the experiences of previous studies of urban sprawl in different cities in the world (see Table 1). Therefore, this study aims to identify the characteristics of urban sprawl patterns in the Batticaloa municipal council using Geographic Information System (GIS) and remote sensing technology. The findings can provide knowledge about the characteristics of urban sprawl to understand the sprawling patterns in other cities in Sri Lanka that have not been addressed so far. Apart from the empirical and methodological contributions, the findings of this study, in line with


**Table 1.** Summary of characteristics for urban sprawl based on previous studies.

measures to control sprawl.

Sustainable Development Goal 11 and the New Urban Agenda, offer useful insights and

#### **2. Definition and Characteristics of Urban Sprawl**

Although urban sprawl is still considered an elusive concept, it has been used around the world for almost eighty years [29]. It was first realised through the transformation of agricultural and forestry areas into industrial, residential, and commercial development in the United States in the late 1950s. The term "urban sprawl" appeared in printed documents in 1960 [30] and is used in various fields, such as urban studies, remote sensing, and geography [18].

Urban sprawl has been defined by its characteristics identified in a particular urban area. Urban sprawl is the encroachment of non-urban lands to urban lands that occurred beyond the built-up area with a leapfrog pattern, unorganised pattern, low-density pattern, and unordered development [9]. Some patterns, such as ribbon development, low-density development, and leapfrog growth, were identified in Aligarh city, India, called urban sprawl [13]. However, urban sprawl was defined by eight metrics for land use, such as concentration, clustering, proximity, mixed-use, nuclearity, density, clustering, and centrality [19]. Sprawl is the irregular urban form with different scales of land use that consist of common and institutional facilities [31].

In contrast, similar patterns, such as uncoordinated, uncontrolled, and unplanned growth, were found along highways, called urban sprawl [20,32]. In addition, urban sprawl is unplanned discontinuous growth categorised by low-density growth in urban boundaries [5]. It is a typical pattern with scattered growth [11,21,22]. Urban sprawl is car-dependent, and low-density development has several negative impacts [22]. Lowdensity patterns and stripe development along major highways were identified in Colombo metropolitan area, Sri Lanka [4]. Sprawl development was identified in three directions along three main roads in Kandy city, Sri Lanka [8], which is almost the similar pattern of Colombo city.

Thus, the definition of urban sprawl varies among researchers who define it based on the characteristics of their urban area [5,7,19,22–24]. Although specialists and researchers still have problems defining the term "urban sprawl" [15,33]. However, many researchers accepted Ewing's (1997) definition is more suitable for recognising the urban sprawling, which stated an urban land use existence, characterised by scattered development, lowdensity, leapfrog development, and commercial strip development [12,34]. According to Table 1, more than 17 researchers have used more than one of these characteristics to explain urban sprawl.

Based on the review (see Table 1), urban sprawl characteristics, which are low density, leapfrog development, commercial strip or ribbon development, scattered and dispersed development, auto-dependent or car-dependent, uncontrolled growth, uncoordinated growth, and unplanned growth, were mainly identified in the developed and developing countries. Some characteristics, such as low density, leapfrog development, and scattered development (see Figure 1), were found in many countries [12,18,21], for example, Indonesia and India. As a developing country in the world, Sri Lanka has many development plans implemented due to the rapid urban growth in recent decades, which is causing urban sprawl.

**Figure 1.** (**a**) Low-density development [35], (**b**) Leapfrog development [35], (**c**) Commercial ribbon development [35] and (**d**) Scattered development [36].

#### **3. Materials and Methods**

#### *3.1. Description of Study Area*

The study area, Batticaloa Municipal Council, is the local authority in the Batticaloa district located in the eastern part of Sri Lanka. The average elevation of Batticaloa is 8.523 m above Mean Sea Level. The total population in this area is 93,306 people. The propagation of ethnic population is as a "sandwich pattern" with Tamil (91%), Muslim (5%), Sinhalese (0.14%), and others (3.86%) [37,38].

The total land area of the Batticaloa Municipality is 4311.87 hectares, which is allocated for different purposes of utilisation such as residential, agricultural lands, commercial, wetlands, water bodies, scrub forests, and others (refer to Table 2). Five (5) land parcels separate this city from the inland water bodies. These land parcels are connected to the bridges for transportation. As a clustered city, the land area uses for multiple purposes and these links with various sectors, such as fishing, agriculture, small industries, and commercial. Each land parcel has different property uses and has cluster development in each sector, such as commercial, recreational, and residential. The most dominant land use of this area is residential (1170.24 hectares), and the next is agricultural lands (935.6 hectares). One of the low proportions of the land-use class is commercial (23.5 hectares), which compares to other major land uses. The natural lands, including wetlands (82.5 hectares), water bodies (58 hectares), and scrub forests (185.71 hectares), are also a certain portion in this area [37].


**Table 2.** Major land use categories in Batticaloa Municipality.

Considering the infrastructure facilities in the Batticaloa municipal area, it is undeniable that ongoing development projects will improve their current conditions. However, these development activities are not planned by urban planners and relevant development officials. Arbitrary development occurs highly in this area which affects the pattern of sustainable land use. Since 1990, the rapid growth of the population has caused several changes in the built pattern that were not assessed by the authorities for sustainable development. Permanent and temporary migration to Batticaloa municipality increased from other parts of the district and the Eastern province due to the effects of the Civil War because this area is the major urban centre in the Eastern Province, Sri Lanka, with all amenities. In addition, the proposed development plan in Batticaloa municipality for 2030 contained many rules and regulations on land use, especially built-up development, which is not considered much further in current development. Thus, Batticaloa municipality began to face the urban sprawl development in the core city and the periphery. As rapidly developing cities in Sri Lanka, Batticaloa received more concern from planners as the city was expected to become a metro city by 2030. Therefore, this area has been chosen to study the urban sprawl in Sri Lanka, which is more significant to understand the characteristics of sprawling. Figure 2 shows the Batticaloa Municipal Council area, the study area in this research.

#### *3.2. Data Collection—Source of Data*

Understanding the characteristics of urban sprawl requires the pattern of land use, especially the built-up changes in the area. Google Earth Pro and ArcGIS 10.6.1 applications were utilised to produce various layers, such as built-up maps, density maps using satellite images, boundary map for Batticaloa municipal council, and Grama Niladhari division map for Batticaloa Municipality.

**Figure 2.** The Study Area—Land use pattern in the Batticaloa Municipal Council—2020. Source: Modified from the Batticaloa Municipal Council Profile, 2021.

There are two (2) types of data, namely, remote sensing data (satellite images) and demographic data, used to generate the maps. Satellite images were downloaded from the Earth Explorer, United States Geological Survey [39]. Details of this data presented in Table 3 with all the information. These time-series images for 2000, 2010, and 2020 used to produce the land use maps to identify the sprawling and extract the built-up area. Meanwhile, the demographic data were collected from the Department of Census and Statistics, Sri Lanka. This data was utilised to produce the population density map in order to identify the density changes. Then, census data for the years 2001, 2012, and 2019, which are the most recent years with satellite images, were compared with the built-up images to understand the influence of population growth on urban sprawl.


Source: Earth Explorer, 2021.

#### *3.3. Data Processing and Analysis*

The downloaded satellite imageries were geo-referenced in World Geodetic System 84 (WGS84) and then projected to the Kandawala local coordinate system. Filters, brightness, and contrast tools were used to improve the quality of the satellite images. The Batticaloa Municipality's boundary was digitised as a shapefile using the current map to demarcate the study area. Based on this boundary, three (3) satellite images, which are from the years 2000, 2010, and 2020, were clipped to delineate the satellite images based on Batticaloa Municipality's boundary. Then, these images were classified into six (6) classes, namely, built-up, agriculture, forest, water bodies, vacant land, and others, according to the training land samples using Supervised Maximum Likelihood classification in ArcGIS. Then, the classified images were validated using the accuracy assessment technique. About 85% of overall accuracy is usually considered enough in the map data [25]. The overall accuracy of land use can be obtained by the following Equation (1) [40]:

$$OA = (1/N)\sum\_{i=1}^{r} n\_{ii} \tag{1}$$

where *OA* is overall accuracy, *nii* is correctly classified pixels' number, N is pixels' total number, and r is rows' number.

Ground truth data was considered in comparing the classified Landsat images. Overall, 127 training samples for ground truth were obtained as random points at specific locations using the grid layout in Google Earth Pro and using known coordinate points. Each point has valid land-use class values, which are:


The confusion matrix was formulated to find the accuracy and obtain individual accuracy between the classified classes and the reference data, such as coordinate points collected from the field. The classification accuracy was determined to obtain the level of precision. The final output of this process was a land use map of Batticaloa Municipality from the years 2000, 2010, and 2020.

Then, density mapping was used, which is a method to show the location of points or lines which can be concentrated in a given area. Such maps often use interpolation methods to estimate a given surface where the concentration of a given function can be. The population density is 1372 persons per km<sup>2</sup> in Batticaloa [37], which is rapidly increasing in recent decades due to rapid urbanisation. Therefore, the population density map was produced using the density analysis tool in ArcGIS. The population for each Grama Niladhari divisions was utilised to produce the density map. The spatial boundary data for Grama Niladhari was developed using the base map of the Batticaloa Municipality. The population data for each Grama Niladhari division were added to the spatial file to show the population's spatial distribution. Based on the standard calculation of the population density, the number of people divided by land area is calculated. The Equation (2) for population density is as follows:

$$PD = \left( T n / a \right) \tag{2}$$

where *PD* is population density, *Tn* is the total population in a particular area, and *a* is the land area in hectares.

It is categorised by suitable data clustering methods, which is Jenks's natural breaks classification, designed to determine the best arrangement of values. It is used to classify the data into five (5) categories, such as very high, high, moderate, low, and very low. This density distribution helps to understand the changes clearly as to which area is high and low density. Figure 3 shows the complete methodological framework for identifying urban sprawl characteristics.

**Figure 3.** The methodological framework for identifying urban sprawl characteristics.

Further, the built-up density map was also produced to establish the low-density development. The buildings of the study area were digitised using Google Earth Pro. Each building's features were converted into points using the feature to point tool in ArcGIS. Then, buildings for each Grama Niladhari divisions were clipped, and density analysis for buildings was conducted to generate the built-up density for the Grama Niladhari division in Batticaloa municipality. The density was categorised by Jenks's natural breaks method into five (5) classes as very high, high, moderate, low, and very low. The density changes in the study area were identified by using these maps. Based on the results also, we can clearly understand which area has more sprawl.

#### **4. Results and Discussion**

The dynamics of the built-up area, known as a typical process of urban sprawl, are of particular importance for understanding spatial patterns for sprawl development. The built-up spatial patterns identified the characteristics of the urban sprawl in the Batticaloa Municipal Council.

#### *4.1. Built-up Patterns*

The built-up patterns are presented in the maps of years 2000, 2010, and 2020 (refer to Figure 4) to understand the sprawling characteristics in the Batticaloa municipal council. The built-up area has an extent of 1162 hectares in 2000, which increased to 1439 hectares in 2010. It increased to around 1557 hectares in 2020, showing the increases in the city's built-up pattern (refer to Table 4). Based on this, a rapid increase in built-up growth was identified during the selected periods.

**Table 4.** The extent of built-up area in Batticaloa Municipality.


The accuracy of the built-up pattern is identified by the classified land use maps of the study area. The analysis revealed that the producer accuracy and the user accuracy varied during the selected periods. Producer accuracy refers to the accuracy of the map with how often real features on the ground are displayed correctly. User accuracy refers to the map user accuracy, which indicates how often the land use class on the map is actually present on the ground. User accuracy shows the reliability of the map. According to the calculation, the producer accuracy for the built-up area in 2000 is 89.43%, while the user accuracy is 96.72%; in 2010 the producer accuracy is 94.6%, while the user accuracy is 99.72%; and in 2020, the producer accuracy is 87.71%, while user accuracy is 93.29%, which is an excellent precision for analysis. The overall accuracy of around 85% is considered enough to prove the precision of the map data [25].

**Figure 4.** *Cont*.

**Figure 4.** Built-up patterns in the Batticaloa Municipal Council in (**a**) 2000, (**b**) 2010, and (**c**) 2020. Note: The Grama Niladhari Division (GN) is a subdivision of the Divisional Secretariat in Sri Lanka. A total of 14,022 Grama Niladhari divisions are in charge of 331 Divisional Secretariat divisions in Sri Lanka; of these, 48 of Grama Niladhari divisions are in the Batticaloa municipality.

The Batticaloa municipality area consists of 48 Grama Niladhari divisions. Thiraimadu is one of the divisions that emerged with housing developments after the tsunami disaster. This area is also developing as an administrative zone in the city that encourages people to construct housing, which is one reason for rapid built-up development in this area. In addition, Puliyanthivu Island is the city centre, with densely developed commercial and residential buildings. However, the lower land value in the Navalady and Thiraimadu areas attracts people with low and middle incomes to buy the land and build a house. The main reason for the lower land value is that this area is often affected by disasters, especially flood. Besides this, high land value has been identified in the Puliyanthivu, Oorani, and Thandavanveli areas because these areas are close to the city centre, the highway, and several infrastructure facilities. One of the best policies is Land Value Capture and Taxation, which is beneficial for affordable housing in this city with a lower land value. This system is in place to increase revenue and fix up the downtown buildings in the city. This income can be used to develop housing for people with low incomes. However, the tax system is already implemented in the city, which is not strictly followed annually. Although everyone is aware of the property tax in the city, they sometimes forget to pay the annual renewal tax. The municipality does not remember and observe these activities regularly, which leads to the illegal land formation as well as sprawling development in the city.

Several groups of people own the total land area in the Batticaloa municipality. This land has been distributed around 73.1% to the inhabitants, 9.1% to the government, 4.7% to Batticaloa municipality, and 13.1% of obscure details. One of the principal regulations is that people cannot construct any buildings in the Batticaloa municipality area without obtaining an approved development permit. However, some people carry out illegal construction development, which represents around 13.1% of the total area of Batticaloa municipality, and those do not contain explicit information about the property. These owners build houses or other buildings on other people's land without getting the proper approval from the municipality. These activities increase the most illegal construction in the city. These developments triggered the formation of scattered and leapfrog development in the city, which are the main reasons for the sprawling growth in this city. Therefore, the property documentation system should be appropriately maintained by the municipality. A survey for property owners should be conducted at the turn of the year to identify illegal land and minimise sprawling growth. This survey can also inspire people to pay taxes without fail. In addition, the municipality should gently remind people before the annual tax period ends. For this, a smart application should be developed for the municipality to identify the pending cases quickly. These practices can control land occupancy and the maintenance of more than one piece of land of a person in the city.

In addition, land ownership problems related to communities in the Batticaloa municipality were identified in the Nochchimunai and Sinna Oorani Grama Niladhari divisions. At the same time, the Puliyanthivu South, Kallady uppodai, Kokkuvil, Punnaicholai, Karuvappankeny Amirthakali, and Mamangam areas identified landowners' issues associated with low-income people. These types of problems caused to form vacant land and more subdivided lands in the city. In addition, illegally divided lands were identified mainly for sale in the Saththurukondan and Uppukarachai area, where the land value is relatively high today. These activities mainly lead to sprawling growth in the city. Therefore, the municipality must circulate a pre-approval method to divide the land in the city. People should inform the municipality about the subdivision of the land, and then the municipal official should visit the specific area to observe the land. After that, the municipal guidelines must be adapted to that particular area, and the entire previous land document must be verified to confirm the land entitlement for subdivision and sale. This practice can find illegal subdividing of land and sale, minimise future land problems within communities, and reduce the amount of vacant land in the city.

Figure 5 shows the built-up changes between the years 2000 and 2010, and between 2010 and 2020 to understand the expansion. This comparison showed a gradual increase in built-up changes in the study area. The built-up growth increased 227 hectares between

2000 and 2010, and 118 hectares between 2010 and 2020 (see Table 5). These gradual changes in the built-up area have illustrated the growth of sprawling in the city. The rapid growth was registered between the years 2000 and 2010. The Batticaloa area was one of the severely affected areas by the civil war in Sri Lanka. Thus, most people from the other parts (rural areas) of the Batticaloa district, such as Porativu, Vellaveli, Mandur, Thikkodai, and Vaharai, migrated to the Batticaloa city for survival, including security, education, and livelihood. Furthermore, the living standards, access to more facilities, and admired city life are the reasons for the migration of people to the city. The movement to the city led to the demand for land and housing for the people. The desire for their own property has created a haphazard development in the city and the urban fringe of the Batticaloa area. Thus, different characteristics were identified in the different areas of Batticaloa city. However, many urban sprawl characteristics exist in the world's cities, but the study area is occupied with some characteristics.

**Table 5.** The changes of built-up area in Batticaloa Municipality.


**Figure 5.** *Cont*.

#### *4.2. Characteristics of Urban Sprawl*

The built-up area was extracted from the land use map to identify the urban sprawling characteristics. The spatial and temporal built-up patterns reveal that sprawling characteristics identified as low density, leapfrog development, scattered growth, and commercial ribbon development influenced the irregular urban development pattern. Most of these characteristics are identified in the city limits, and some are in the core city, which affects the city's sustainable growth.

#### 4.2.1. Low-Density Development

Low-density development is one of the main phenomena of urban sprawl generally risky to the urban environment. The primary units for identifying the urban sprawl, including density, are buildings, especially housing units of a particular area [19]. The residential developments are mainly identified in the marginal low-density areas in the city. Low-density development is a piecemeal extension of the built-up area, which consumes much land in the urban fringe. It is the most generally indicated characteristic of urban sprawl in many pieces of literature [41]. Residential housing mostly consumes the vast land, which was vacant land previously, leading to the low density. The rise of land and property value in the city cannot afford a vast population; however, this value is meagre in the urban fringe. Thus, the sprawl areas are occupied mostly by the low-income people for their permanent residence. They are attracted by these vast, spacious living areas to build an affordable house [24], also experienced by the Batticaloa area. The housing preference of the lower class and some middle-class people pushed them to settle in these low land

value areas. The people who migrate from the village areas admired the city limits, which is more spacious and affordable for their own housing. In addition, a single dwelling unit in a larger area in the Batticaloa municipality is one of the main reasons for the low-density development, which is similarly identified in the United States of America [24].

Figure 6 shows the built-up density in the Batticaloa municipal council area by Grama Niladhari division. Based on this, higher density patterns were identified in the city centre from 2000 to 2020. However, nine divisions, which are Saththurukondan, Thiraimadu, Paalameenmadu, Panichalady, Kokkuvil, Thiruperunthurai, Thimilathivu, Veechikalmunai, and Navalady, mainly come under the low density in the divisions throughout these periods. The lack of space in the city centre for housing development has been limited to low-income residents where land value is most in demand. Further, nearly 150 low-income families living in the city had the rural characteristics identified in Sinna Oorani, Punnaicholai, Thiraimadu, Mamangam, Kokkuvil, and Saththurukondan areas. The developed built-in density map can be useful for identifying areas with low and very low density in the city. Based on this, an appropriate development plan can be adapted to this area to make the city more compact.

Table 6 shows the range of built-up density in the Batticaloa Municipality. The ranges were categorised from very low density to very high density. Based on this, around 0–4 buildings per hectare were identified in the very low-density areas and more than 22 buildings per hectare in the very high-density areas. The built-up density increased near the city centres, except in the northern and western part of the city, from 2000 to 2020. The main reason for the low density in these areas is the inadequate facilities such as accessibility to highways, commercial, and other services.

**Figure 6.** *Cont*.

**Figure 6.** Built-up Density in the Batticaloa Municipal Council in (**a**) 2000, (**b**) 2010, and (**c**) 2020.


**Table 6.** The range of built-up density in Batticaloa Municipality.

Most of the lands are used for single use, like individual housing, which created the low-density development in the Batticaloa municipality. For example, 88% of homes are single-storey separated houses, 9% of homes are two-storey separated houses, and 1% of homes are more than two-storey separated houses. These housing patterns show a rural characteristic in this city. In addition, these single housing developments are one of the main reasons for the low-density development in this area. Therefore, housing policy must be designed in accordance with Sustainable Development Goal 11 and the existing situation of Batticaloa municipality. Housing policy should be developed in consultation with stakeholders in Batticaloa Municipality who provide a clear view of all income earners and the different communities living in the city. The municipality then displays the decision for public responses that provide different ideas for improving the policy before it is implemented. Furthermore, reporting the municipality's policy in public can also make the right decisions in all activities by people, including building houses and maintaining the land.

In addition, building codes must be provided to track building types and the location of buildings in the city. This method can help to quickly identify a specific building in all situations and demolish illegal constructions. These activities primarily help control the future sprawling growth of the city. Further, the developed built-up pattern and density maps are helpful to identify the additional unregistered buildings in each Grama Niladhari division. For example, a Grama Niladhari division already has 25 buildings registered in the municipality, but the map shows 28 buildings in the same division. By this, the constructions can be understood as illegal development in the area in question. The municipality can take the necessary measures against them and also minimise the sprawling growth in the city. In addition, a monitoring unit should be set up to review housing policy and the necessary strategies for building construction in the city. An online platform should be developed to guide public discussions and consultations. This continuous monitoring activity can control illegal housing development in the city.

The population density is measured by the ratio of people inhabiting a specific region in persons per square kilometre or hectare. A city that occupies a smaller land area is considered more compact and less sprawled, and that with more extensive land occupied by less population implies low density and a more sprawled characteristic [41]. Table 7 shows the range of population density in the Batticaloa Municipality. Population density ranges from very low to very high density. Areas with 0 to 10 persons per hectare are known as very low-density areas. Areas with over 100 people per hectare were identified as very high-density areas in 2001, and with more than 84 people per hectare were very high-density areas in 2012 and 2019. The population density is almost high in the city centre and close to commercial areas.


**Table 7.** The range of population density in Batticaloa Municipality.

The consumption of the land is faster than the population growth, which revealed low-density development. As shown in Figure 7, the population density in the Batticaloa Municipality is identified by the Grama Niladhari divisions. Based on this, developed areas such as Puliyanthivu, Arasady, Thandavanveli, Oorani, Kallay, Iruthayapuram, Mamangam, Thiruchendhoor, and Navakkudah are almost high built-up density areas. However, those areas have not populated much when compared with built-up. However, the highest population density was identified in the Arasady, Koolavady and Iruthayapuram areas. Low density was registered in the edge areas, but the core city and the highway area only showed higher population density and built-up. Paalameenmadu, Navalady, Saththurukondan, Puliyanthivu west, and Thiruperunthurai areas were identified with the lowest population density patterns in the Batticaloa municipality. The main reason for the low density is poor accessibility to highway and other services. In addition, Seththukudah, Vechukalmunai, and Thimilaithivu areas were under military control during the civil war period. Thus, people did not desire to make their settlement in those areas.

**Figure 7.** *Cont*.

**Figure 7.** Population Density by Grama Niladhari Divisions in (**a**) 2001, (**b**) 2012, and (**c**) 2019.

Additionally, Batticaloa Domestic Airport is located in the Thirupperunthurai division, which is closest to the Puthunagar and Sethukkudah areas. Residential buildings are banned around the airport areas such as Puthunagar, Thirupperunthurai, and Sethukudah due to the airport expansion project. In addition, a water supply system was launched in the high-density populated areas such as the city centre and the nearby areas. In other areas, people use well water for their needs. Further, 38 schools are located in the Batticaloa municipality area. Of these, seven schools are national schools with a high level of education, located in the city centre and the nearest areas, which is one of the reasons for the high population density.

Further, the population of this city has a high growth rate of 3.92% in the period 1990–2001 and around 2.07% in the period 2001–2010. This rate changes to around 2.27% in the period 2010–2019, which indicates the fastest growth of cities in Sri Lanka. In addition, the population growth rate in the future is expected to be 2.5% to 3.0% in 2030. The minimum expected population is 127,291 persons and a maximum of 170,714 people in 2030 [37]. This rapid population growth can lead to more sprawling development when political influences disrupt municipality development plans and regulations. Therefore, the rules and regulations must be strictly followed to become a sustainable city in the future.

#### 4.2.2. Leapfrog Development

A discontinuous irregular pattern on developed land is widely recognised within the city limits. This type of development makes it costly to provide essential services like water and drainage. This development consumed a wide range of land and created an arbitrary development pattern that destroys urban beauty. Figure 8 shows the leapfrog development, which creates more changes in the land use pattern, leading to the urban sprawl established in the visual map. This development can identify a very inefficient land use pattern, which is one of the most extreme examples of urban sprawl. Such growth affects the development of the city directly, including infrastructure and services.

The leapfrog development creates less housing and population density due to the undeveloped land, such as the urban fringe. This density is higher than the individual homes, which affects sustainable development [21]. The vacant land in specific areas such as Thiraimadu, Paalameenmadu, Kokkuvil, Panichalady, and Navalady are good examples of leapfrog growth with low density in the Batticaloa municipality area. Fundamental accessibilities such as public transportation, telecommunication, and water supply are comparatively poor, leading to less population growth. These people live in small housing units, which means one-room or two-room houses built using bricks or clay, but the land extent is larger than the houses. The people living in these areas are from low-income classes. The municipality should introduce more housing schemes, incorporating with the National Housing Development Authority, Sri Lanka. This development can reduce the leapfrog development within the city limits. In addition, the municipality should raise people's awareness about the leapfrog development and its impacts on the city through community programmes. People's understanding of this issue can regulate builtup development.

The leapfrog pattern develops due to the spatial heterogeneity of agricultural facilities [42], which is the case in the Batticaloa municipality area, such as Thiruperunthurai and Puthunagar. Thus, this discontinuous pattern forms the vacant land between built-up areas, making it difficult to afford facilities by the municipality. The transformation of non-urban land influences leapfrog and edge development into built-up land that increases faster than the growth of population in the Chinese cities [9]. However, this transformation was not identified in the Batticaloa municipality area during the selected periods. Rather, built-up discontinuous patterns were mainly identified in this city, which cause the rising cost of infrastructure.

**Figure 8.** A leapfrog development pattern in the Batticaloa MC—2020.

#### 4.2.3. Scattered Development

Scattered development also provides an inaccessible pattern on the urban edge, like undeveloped areas, creating sprawling. Figure 9 illustrates the scattered patterns in the study area that grows in the urban limit. The built-up growth develops in a dispersed way, which creates a considerable change in the city. This pattern was mainly identified in Thiruperunthurai and Navalady areas in the Batticaloa municipality. One reason for the scattered growth in the Thiruperunthurai area is that the land is mainly used for agriculture purposes. Municipal open spaces assigning a convenience value is one of the ways to account for scattered development. Therefore, people can be willing to spend more money on owning a home in these spacious areas, even if they are further away from the central business district, such as in the European cities [42]. This activity is sometimes experienced in Batticaloa city as well, in recent decades.

Further, the settlement areas of the core city developed by the individual homes in a vast land are mostly composed of single-storey buildings and some of double-storey houses. The individual housing preference of these people induces them to occupy the spacious land for building their dream house. Most families here are nuclear families rather than the extended family needed to build many individual houses for each family. For example, a nuclear family has four members, such as a father, mother, and two daughters. The parents must build two houses for these two daughters to marry them. Parents must build an individual house for each daughter, even if they have five daughters, because of the tradition in Batticaloa. A better solution for this, parents can build a low-rise building like three or four floors with all facilities and assign each floor to each daughter, which

can minimise the number of individual single homes in the city. This practice should be included in municipality policy, which makes the city more compact.

**Figure 9.** A scattered development pattern in the Batticaloa MC—2020, (**i**) and (**ii**) shows the scattered pattern in some GN divisions in the city.

#### 4.2.4. Commercial Strip or Ribbon Development

Commercial development, along with highways, is another characteristic of urban sprawl called ribbon development that threatens sustainable urban growth in the city. Commercial buildings are mostly built along the main transport corridors in the core city and outside of the downtown area (see Figure 10). This type of development caused an increase in the value of the land near the highways. These urbanised areas use a mixed mode, such as commercial and residential, affecting urban land use. Many buildings utilise only the ground floor for commercial purposes, though in some buildings the second and third floors are also used as commercial spaces. Nevertheless, in the rest, all second or third levels are mainly used for residential purposes. People want to buy things such as clothes, grocery items, and electronic products for different purposes simultaneously, and they have to walk long distances to buy these things. This distance affects people's continuous shopping and time consumption.

**Figure 10.** Commercial Ribbon Development in the Batticaloa MC—2020.

Further, this commercial ribbon pattern causes several problems in the city's functions. For example, approximately 75,000 to 100,000 people commute to Batticaloa municipality from 6 a.m. to 8 p.m. every day. The reasons for this commuting are to access services such as the railway station, the teaching hospital, the Faculty of Healthcare Sciences at the Eastern University, the Swami Vipulananda Institute of Aesthetic Studies at the Eastern University, the Open University, the district court, the airport, the technical college, the financial institutions, and other government institutions. Most of these institutions are located in commercial areas. Thus, this continuous commuting activity causes traffic congestion and frequent accidents in the city. Mainly Koddamaunai Bridge and the new bridge areas face traffic congestion daily. The main reason is that the traffic lights are not fixed in many areas of this city. Further, about 500 private buses and 320 public buses operate in the Batticaloa municipality area. This bus service starts from the Batticaloa's central bus station, which is located in the city centre, and has caused overcrowding. This ribbon development pattern is one of the main reasons for congestion in the city.

However, all these characteristics have created similar and different effects on the Batticaloa municipality. Leapfrog and scattered developments are the discontinuous and dispersed built-up growth that have less connectivity between the buildings. Meanwhile, commercial ribbon development with shopping complexes, restaurants, and banks built along the street of the core city generally depends on the highways for the developments. Various development projects were implemented in this area at the end of the civil war, which caused more changes in the city. Thus, land use classes were analysed to extract the area through characteristic changes in urban sprawl during the years 2000, 2010, and 2020.

The building patterns are not growing in a planned manner, such as housing, administration, shopping complexes, schools, police station, health office, and playgrounds. The police station and its quarters are built between the commercial area and also the core city. At the same time, administration buildings and playgrounds are also built in the core city. The suitability for the development of each sector was not considered until now. The main reason is less development planning in this area because it was affected by the civil war for three decades (1983–2009).

Further, the transportation network in the city was also not well designed. The primary and minor roads were not planned based on the network tracking method, which is more familiar for making road maps in ArcGIS. The road network of Batticaloa is not well mapped to identify the closest route between two locations in order to avoid traffic-related problems. For example, it is difficult to reach the general hospital of Batticaloa because of the lack of the closest facility in an emergency during traffic congestion. Closest facilities must consider tracking the places as quickly as possible to reach the location, and this must be a consideration when developing the transport pattern.

As a developing city in Sri Lanka, in Batticaloa the remaining categories that cause sprawl, such as auto-dependent or car-dependent development, uncontrolled growth, and uncoordinated growth, are not highly identified in this study area. The lower population of the city is the reason for these characteristics not having grown during the selected periods. Therefore, based on the identified characteristics of the study area, urban sprawl refers to the urban expansion [9,43] with low density beyond the built-up area [1,5,9,12,22,23], leapfrog development [5,12,13,26], a commercial ribbon development along the highways [12,13], and scattered growth [11,22,25].

#### *4.3. Future Plans and Regulations in the Batticaloa Municipality*

Plans and regulations regarding the construction of buildings, especially the construction of homes in current and future development, are the most important to minimise the urban sprawl development in the future. The development plan for Batticaloa municipality in 2030 is proposed in nine major zones, which are: residential zone, commercial zone, information technology zone, environmental conservation zone, airport-related activity zone, mixed development zone, administrative zone, fort conservation zone, and agricultural zone. These zones are separated by boundary lines such as roads or railways, or canals. These zones will only be used for the specific development for which the areas were designated. However, Batticaloa as a tourist city encourages activities related to tourism in any area of the municipality, depending on the suitability of the tourist development.

Further, the minimum land extent for the residential building is 6 perches, mentioned in the regulations of Batticaloa municipality. However, the maximum land area is not defined. Therefore, there are no obstacles to buying a large land and building a single house on this large land. Some people prefer the large spacious land to build their home today. The land for their preference is mostly available in the peripheral areas of the city. This development is the reason for the low density and scattered development patterns in the future as well.

Further, all buildings within 300 m of the coastal zone must be constructed with the prior approval of the Department of Coastal Conservation, Sri Lanka. A green belt was developed in the coastal strips to control the loss of biodiversity and the barriers from disasters such as tsunami and cyclone. Any building constructions in proximity to these disaster-prone areas should obtain clearance from the Urban Development Authority, Batticaloa. These regulations are the reasons for the low population and built-up density in the coastal areas of the city.

In addition, environmentally sensitive areas such as scrubland and mangrove forest are considered conservation areas, which do not allow the built-up development in the municipality areas, showing low density and leapfrog characteristics. Even though people can get approval to build houses in some environmental conservation areas, the buildings should be designed to retain natural beauty rather than obstruct the open green spaces. Therefore, people should follow the regulations mentioned by the municipality to control the urban sprawl development.

However, Land Value Capture and Taxation are favourable for affordable housing development in the city with lower land value. Proper activation of this system can increase revenues and fix up downtown buildings. The significant increase in land value is due to the locational preferences as commercial areas and a transport hub. This tax system can control the increase of land value by the community preferences and use these tax revenues to develop public spaces such as bus stands, roads, and public spaces. Thus, the Municipality should strictly follow this policy in this city and encourage people to adopt this system to control land value, since people who pay tax can force the authorities to take care of their land and control the illegal occupancy of land in the city.

In addition, the urban development plan must also be open to feedback from stakeholders and experts who can make the development plan stronger. Most of the time, the municipality does not follow this practice when trying to implement a policy. The public is largely ignored in this process, which affects the sustainable planning of the city. Thus, the municipality should consider development activities in all aspects, including public participation. Therefore, proper regulations should be implemented based on the Sustainable Development Goal 11 and New Urban Agenda.

#### **5. Conclusions**

Urban sprawl has been increasing due to rapid construction development, especially housing in the Batticaloa municipal council. Urban fringe in the study area showed different sprawling characteristics, which are low density, leapfrog development, and scattered growth. However, a characteristic such as commercial ribbon development along the highways has been identified in the core city areas. Rapid population growth causes more sprawling development in the city, which was increasingly identified after 1990 in this area. In addition, people's preferences, land value, education, employment, income level, illegal construction, and permanent and temporary migrations are the main reasons for the sprawling development of the city. Based on these factors, it is clear that urban sprawl is a socioeconomic phenomenon that should be focused more on socioeconomic aspects in future studies.

Further, political influence interrupts the municipality's development plans and regulations. The rules and regulations need to be followed carefully in this area to develop itself as a sustainable city in the future. Maximum land extent for housing development should be stated in the regulations, which are the most important to minimise the pattern of low density and scattered development in the city. The municipality must identify existing illegal buildings, and when the buildings have not been built following the regulations, they must be demolished by the municipality. In addition, the illegal landfill in wetland areas such as scrubland areas must be punished and charged a penalty by the municipality. These regulations can control illegal activities in future. People have a lack of awareness of the urban sprawl development in this city; thus, it is important to educate them about the urban sprawl development and its effects through research, mapping, and community programmes.

Finally, this study identified the particular characteristics of urban sprawl such as low density, leapfrog development, scattered growth, and commercial ribbon development in the Batticaloa municipality. This finding empirically contributes to understanding the patterns of uncontrolled urban sprawl in Batticaloa city and other cities of Sri Lanka, and other developing countries in the future. This study can help to formulate strategic policies to minimise sprawling growth in the Batticaloa Municipal Council. Despite the above, this study, nevertheless, has limitations. Although the current low-resolution satellite images are deemed adequate in terms of accuracy, the findings could be enriched via a more highresolution images analysis to identify micro-level changes and built-up patterns/forms. That may, in turn, provide more interesting explanations about urban sprawl characteristics. In addition, these sprawl characteristics creating impacts to land use/land cover patterns

are associated with physical and socioeconomic influences, but they are not considered in this study. Therefore, future studies should consider these influencing factors to obtain a holistic understanding and then form a predictive model curbing urban sprawl more effectively.

**Author Contributions:** Conceptualisation, methodology, M.S. and N.R.; software, M.S.; validation, M.S., N.R., G.H.T.L. and I.S.; formal analysis, M.S.; writing—original draft preparation, M.S.; writing—review and editing, M.S., N.R., G.H.T.L. and I.S.; supervision, N.R., G.H.T.L. and I.S.; funding acquisition, N.R. and G.H.T.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Acknowledgments:** We are very grateful to the Center for Innovative Planning and Development (CIPD) and the Faculty of Built Environment and Surveying, Universiti Teknologi Malaysia for the encouragement and financial support for this study, and our gratitude also goes to the reviewers who provided constructive comments and insights on this manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **Monitoring and Modeling the Patterns and Trends of Urban Growth Using Urban Sprawl Matrix and CA-Markov Model: A Case Study of Karachi, Pakistan**

**Muhammad Fahad Baqa <sup>1</sup> , Fang Chen <sup>1</sup> , Linlin Lu 1,\* , Salman Qureshi <sup>2</sup> , Aqil Tariq <sup>3</sup> , Siyuan Wang <sup>4</sup> , Linhai Jing <sup>1</sup> , Salma Hamza <sup>5</sup> and Qingting Li <sup>6</sup>**


**Abstract:** Understanding the spatial growth of cities is crucial for proactive planning and sustainable urbanization. The largest and most densely inhabited megapolis of Pakistan, Karachi, has experienced massive spatial growth not only in the core areas of the city, but also in the city's suburbs and outskirts over the past decades. In this study, the land use/land cover (LULC) in Karachi was classified using Landsat data and the random forest algorithm from the Google Earth Engine cloud platform for the years 1990, 2000, 2010, and 2020. Land use/land cover classification maps as well as an urban sprawl matrix technique were used to analyze the geographical patterns and trends of urban sprawl. Six urban classes, namely, the primary urban core, secondary urban core, sub-urban fringe, scatter settlement, urban open space, and non-urban area, were determined for the exploration of urban landscape changes. Future scenarios of LULC for 2030 were predicted using a CA–Markov model. The study found that the built-up area had expanded in a considerably unpredictable manner, primarily at the expense of agricultural land. The increase in mangroves and grassland and shrub land proved the effectiveness of afforestation programs in improving vegetation coverage in the study area. The investigation of urban landscape alteration revealed that the primary urban core expanded from the core districts, namely, the Central, South, and East districts, and a new urban secondary core emerged in Malir in 2020. The CA–Markov model showed that the total urban built-up area could potentially increase from 584.78 km<sup>2</sup> in 2020 to 652.59 km<sup>2</sup> in 2030. The integrated method combining remote sensing, GIS, and an urban sprawl matrix has proven invaluable for the investigation of urban sprawl in a rapidly growing city.

**Keywords:** urban sprawl; Landsat; CA–Markov model; SDG 11; urban sustainable development

#### **1. Introduction**

Urbanization is a complex socioeconomic process that shifts the distribution of a population from dispersed rural settlements to dense urban settlements [1]. In spatial terms, the urbanization process is manifested in the physical development of urban settlements and the transition of landscapes into urban forms [2,3]. In the Global South, rapid and unplanned urban sprawl leads to problems such as fragmented landscape, reduction in arable

**Citation:** Baqa, M.F.; Chen, F.; Lu, L.; Qureshi, S.; Tariq, A.; Wang, S.; Jing, L.; Hamza, S.; Li, Q. Monitoring and Modeling the Patterns and Trends of Urban Growth Using Urban Sprawl Matrix and CA-Markov Model: A Case Study of Karachi, Pakistan. *Land* **2021**, *10*, 700. https://doi.org/ 10.3390/land10070700

Academic Editors: Luca Salvati, Iwona Cie´slak and Andrzej Biłozor

Received: 17 June 2021 Accepted: 1 July 2021 Published: 2 July 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

land, increase in urban poverty, and environmental degradation, which pose a huge threat to sustainable development in these regions [4–6]. By 2030, Sustainable Development Goal 11 of the United Nations intends to make cities and human settlements more inclusive, safe, resilient, and sustainable [7]. Building policies to promote the sustainable development of cities, especially in developing countries, need accurate and timely monitoring and understanding of the spatial growth of urban settlements [8].

Geospatial techniques have enabled the analysis and forecasting of urban growth at regional and global scales. These methods are useful for observing and understanding the dynamics of urban landscapes [6,9,10]. Previously, efforts have been made to model and analyze urban spatial growth and patterns using methods such as cellular automata [11–13], the artificial neural network [14,15], the Markov chain [16,17], geographical weighted regression [18], the non-ordinal and Sleuth model [19–21], the analytic hierarchy process [22], machine learning models [23,24], and an urban sprawl matrix [25,26]. Batty demonstrated how cellular and agent-based models have the ability to clearly incorporate spatial interaction and mobility [27]. Considering the limitation of basic logistic regression models, Arsanjani et al. used a hybrid model to uncover the interaction of various environmental and socioeconomic variables that cause urban expansion [28]. By combining the CA model's benefit of modeling spatial variation in complex systems with the Markov model's advantage of long-term prediction, the CA–Markov model was developed, which is an effective method for simulating LULC transformation. It has been widely applied to examining and measuring urbanization and landscape dynamics [29]. The Markov model predicts the future status of a land use based on its current rate [30]. Cellular automata (CA) detects the geographic location of changes, whereas the Markov chain predicts future change based on the past [30].

Karachi, Pakistan's largest city, has seen massive urban growth in recent decades not only in the city's center, but also in the surrounding suburbs [6]. If the urban land expansion rate is higher than the population increase rate, the population density in the urban area will significantly decline, and the phenomena of urban sprawl will occur. Due to institutional inefficiency and governance failure, rural lands have been converted into residential and industrial areas without considering the urban planning schemes in Karachi [31]. The massive conversion of rural lands for urban areas has caused the sprawl phenomenon since 2000, which has led to loss of agricultural lands, an increase in commuting costs, and flooding [31]. The unplanned urban sprawl has also resulted in a range of social problems such as a lack of health care, shortage of education facilities and infrastructures, an increase in criminal incidents, and sociocultural imbalance [32–34]. The introduction of new urban forms and structures that adapt to climate change issues can mitigate the environmental problems caused by dispersed urban area growth and create more efficient urban economies [35]. Therefore, the spatiotemporal modeling of urban sprawl is crucial to better understand the changing urban patterns of Karachi divisions, thus helping local governments in prioritizing the demands of the local population and formulating strategies and practical solutions to achieve the goal of urban sustainable development.

Previous studies have attempted to use remote sensing data to analyze the general pattern of urban land cover changes and urban suitability in Karachi [36,37]. Although land use land cover changes were significant based on the analysis using satellite imagery, the landscape changes during the urbanization process were not fully investigated. Moreover, the simulation and prediction of future LULC scenarios in the growing city have barely been reported. To fill such gaps, this study aimed to thoroughly analyze the LULC changes and the spatiotemporal dynamics of urban expansion in Karachi using satellite data from 1990 to 2020. The future LULC scenarios and urban expansion were also simulated using a CA–Markov model in the city for the year 2030.

#### **2. Study Area and Datasets**

#### *2.1. Study Area*

Karachi, the provincial capital of Sindh, is Pakistan's largest and most densely populated megacity. It is the principal industrial center, seaport, and financial and commercial hub. Karachi Urban Agglomeration (Karachi UA), extending over 3527 km<sup>2</sup> , is located on the coastline of the Arabian Sea, between 24◦45′ N to 25◦15′ N and 66◦37′ E to 67◦37′ E (Figure 1). Karachi is mainly made up of flat rolling plains with hills on the western and northern boundaries. The southern and southeastern banks of the Malir River have a contagious linear concentration of urban settlements [38].

′ ′ ′ ′

**Figure 1.** Study area. PJ = Punjab; SH = Sindh; BL = Balochistan; IS = Islamabad; AJK = Azad Jammu Kashmir; GB = Gilgit-Baltistan; KPK = Khyber Pakhtunkhwa.

According to the 2017 Census Report [39], more than 16 million people live in Karachi, and the population will increase to more than 20 million by 2025 with a density of 4115 persons per square kilometer [40,41]. The city consists of seven districts, which can be further divided into 31 sub-divisions [39]. As an increasing metropolitan city in a developing country, Karachi faces unplanned urban expansion, inappropriate essential infrastructure and facilities, crises in drinking water and solid-waste management services, inconvenient public transport, environmental pollution, and poor governance [42]. Of the total population, nearly 40% live in slum areas [43,44].

#### *2.2. Datasets*

The primary data source for measuring urban spatial patterns and analyzing the trend of urban growth in Karachi was Landsat Thematic Mapper (TM) pictures from 1990, 2000, and 2010 as well as Landsat 8 OLI images from 2020 from Google Earth Engine (Table 1). The atmospheric correction technique LaSRC was used to correct the available Landsat Surface Reflectance Tier 1 data in Google Earth Engine. The CFMASK algorithm was used to mask cloud, shadow, and water regions in these images. The entire study area covered three Landsat tiles (152\_042, 152\_043, and 153\_043). The atmospherically corrected and cloud removed images with a ten-year interval were used to perform the initial LULC classification. As supplementary features for land cover classification, the

normalized difference vegetation index (NDVI) and the normalized built up index (NDBI) were computed for each decadal image [45].


**Table 1.** Details of datasets used in this study.

Several datasets were used as supplementary data in our study (Table 1). To distinguish LULC classes between plain and hilly areas, SRTM digital elevation model (DEM) data were employed. To evaluate the accuracy of LULC, high spatial resolution images with multiple acquisition dates collected from Google Earth and topographical maps published by the Survey of Pakistan, Government of Pakistan were used as reference data. District-level population data were gathered for the years 1990, 2000, 2010, and 2020 from the official census and Pakistan Bureau of Statistics [39]. The road network data were used to train the CA–Markov model for the LULC scenario simulation.

#### **3. Methods**

The workflow was primarily comprised of three steps: classification of land use/land cover, analysis of urban expansion, and modeling of future LULC scenarios. Figure 2 depicts the entire data processing workflow adopted in this study.

#### *3.1. Land Use/Land Cover Classification*

We used the Google Earth Engine's random forest classification technique to produce land use/land cover maps for the years 1990, 2000, 2010, and 2020 in the study area [46]. The overall accuracy (OA), producers' accuracy (PA), and users' accuracy of the classification results were measured using the confusion matrix [46].

#### *3.2. Urban Landscape Change Analysis*

The post-classification change matrix methodology was used to create a land use/land cover change map from 1990 to 2020. To analyze land use/land cover changes, a transition model was developed using cross-tabulation in the GIS module. The transition matrix indicates that the study area had experienced major alterations.

An urban sprawl matrix was utilized to examine urban expansion dynamics and measure urban spatial patterns in Karachi [47]. For the categorization of urban spatial patterns, matrix functions based on urban pixels were used. Using the urban sprawl matrix, the study area was divided into six classes, namely, the urban primary core, urban secondary core, suburban fringe, scatter settlement, urban open space, and non-urban area (Table 2 and Figure 3).

**Figure 2.** The data processing workflow in this study.



**Figure 3.** Depiction of the function of the urban sprawl matrix and urban spatial pattern. The percentage of the built-up area in a 1000 m radius circle centered on the pixel under examination is the pixel's urbanness.

#### *3.3. LULC Simulation*

CA–Markov simulation was implemented using several steps: (a) the generation of LULC maps with the same time interval (1990, 2000, 2010, and 2020); (b) the calculation of transition probability matrices based on LULC maps; (c) the generation of transition suitability maps using driving factors such as distance to water body, distance to main roads, distance to built-up areas, and slope [4,31]; (d) the evaluation of the model's ability to simulate future changes using a kappa index of agreement (KIA) approach; and (e) the simulation of LULC maps for the predicted year (here, 2030). The projections of LULC change in the study area were performed using the land change modeler (LCM) within the TerrSet software (Clarke Labs 2019, https://clarklabs.org (accessed on 10 December 2020) [48].

As an input to the CA–Markov model, the Markov chain model was employed to produce a transition probability matrix between an initial state and a final state. The transitional probability matrices were generated using LULC information from 2010 to 2020 in order to investigate how each land cover class was expected to change. The Markov model can be described using the following equation:

$$\mathbf{S} \ (\mathbf{t} + \mathbf{1}) = P\_{\vec{l}\vec{l}} \times \mathbf{S}(\mathbf{t}) \tag{1}$$

where S represents the land use condition at time t; S (t + 1) represents the land use status at time t + 1; and *Pij* is the transition probability matrix in a certain state, which is calculated using the following equations [49]:

$$|||P\_{ij}|| = \left\| \begin{array}{cc} P\_{1,1} & P\_{1,2} & P\_{1,N} \\ P\_{2,1} & P\_{2,2} & P\_{2,N} \\ P\_{N,1} & P\_{N,2} & P\_{N,N} \end{array} \right\| \tag{2}$$

$$\{0 \le P\_{\vec{\eta}} \le 1\}\tag{3}$$

where *P* refers to the transition probability; *Pij* refers to the probability of changing from state *i* to state *j* in the next time; and PN refers to the state probability of any time. The low transition probability is close to 0, and the high transition probabilities is close to 1 [49].

Using the multi-criteria evaluation (MCE) module, suitability maps, which show the suitability of cell transformation for a particular land cover type, were created for the application of the CA model. The characteristics of LULC types were taken into consideration. For example, the built-up area cannot be converted into a water body [50,51]. As an inherent part for geospatial modeling, the kappa index of agreement (KIA) representing the model's simulation accuracy was used here to evaluate the model's ability to simulate the spatial pattern of land use [52,53]. The KIA was calculated with the following equation:

$$\text{KIA} = \text{Pr(a)} - \text{Pr(e)}/1 - \text{Pr(e)}\tag{4}$$

where Pr (a) refers to the observed agreement, and Pr(e) refers to chance agreement. The kappa coefficients (K-no, K location, and K-standard as well as the overall kappa coefficient) were used to compare the simulated and the LULC map based on remote sensing data of 2020. The kappa coefficient values were calculated using TerrSet IDRISI software.

#### **4. Results and Discussion**

#### *4.1. LULC Change*

In the study area, six LULC classes were identified: bare land, built-up area, cultivated land, grassland and shrub land, water body, and mangroves (Figure 4). According to the accuracy assessment results, the overall classification accuracies were 89, 91, 91, and 89% for 1990, 2000, 2010, and 2020, respectively. The kappa coefficient values were 0.86, 0.90, 0.89, and 0.87 for 1990, 2000, 2010, and 2020, respectively.

**Figure 4.** LULC map of Karachi in (**a**) 1990, (**b**) 2000, (**c**) 2010, and (**d**) 2020.

− −

− −

−

−

Over the last three decades, the land use/land cover in the study region has changed dramatically (Table 3). Between 1990 and 2020, the area covered by built-up area and grassland and shrub land expanded, while the area occupied by agricultural land, mangroves, and open bare ground declined. Divergent changing trends were revealed in the time periods before 2000 and after 2000 for cultivated land, grassland, and shrub land, and mangroves (Table 3). The increase in the area of mangroves and grassland and shrub land since 2000 indicates that afforestation programs have played a positive role in improving vegetation coverage in the study area. The Sindh Forest Department made great efforts to restore and plant endangered mangrove species. With the help of local communities, they planted more than 800,000 saplings of Rhizophora mucronata mangroves in the coastal zone of Pakistan in 2013 [54]. The decrease in cultivated land was observed near the built-up area, which indicates urban expansion at the cost of cultivated land (Figure 4).

**Table 3.** Areal changes in each land use land cover type in Karachi.


The increase in urban areas in different districts of the study area is illustrated in Figure 5. It was observed that districts near coast and far from the core area (Karachi Central, South, and East districts) had a record high urban growth from 1990 to 2020, particularly in the Malir (417.92%), West (279.38%), and Kiamari (257.05%) districts (Table 4). Among the core areas, the East district of Karachi experienced a higher increase in the built-up area than that in the Central and South districts of Karachi. The central city's congestion caused outgrowth at the periphery of the megacity during the study period. As a main driver of built-up area growth, the density of the population in Karachi has constantly been increasing over the last three decades. The population of central city has remained highly concentrated, and its population increased from 1.8 million in 1990 to 3.09 million in 2020. Simultaneously, the population of the suburban Malir and West districts increased from 0.8 million and 0.7 million in 1990 to 2.8 million and 2.23 million in 2020, respectively [39].

**Figure 5.** Built-up area expansion in Karachi in 1990, 2000, 2010, and 2020.


**Table 4.** Built-up area increases in each district of Karachi (in percentage).

Figure 6 shows the land transformation in various districts and time periods induced by the process of urbanization. The majority of areas converted to urban land at the expense of open bare land, grassland and shrub land, and agricultural land (Table 5).

**Figure 6.** Spatial pattern of land transformation in Karachi during the periods of (**a**) 1990–2000, (**b**) 2000–2010, and (**c**) 2010–2020.

**Table 5.** Land transformation from other LULC classes to built-up area during the periods of 1990–2000, 2000–2010, and 2010–2020.


#### *4.2. Urban Landscape Change*

The urban sprawl matrix was used to create urban landscape maps in the study area. The area of urban primary core increased from 145.9 square kilometers in 1990 to 363.5 square kilometers in 2020 (Table 6). In 1990, changes in the area of the primary core were registered in the areas that comprise the CBD area, namely, the South, East, and Central districts, and later in 2020, the urban primary core expended further into the suburban districts of Karachi such as the Malir, West, and Kiamari districts (Figure 7). The area of the urban secondary core also changed from 25.9 sq.km 1990 to 22.3 sq.km in 2020 (Table 6). In 1990, the urban secondary core was observed only in the districts of Malir and Korangi, while later in 2020, the urban secondary core could be observed in other suburban areas of Karachi such as the districts of West and Kiamari. The observed urban secondary core areas in 1990 merged with the urban primary core in 2020 due to rapid expansion, and a new urban secondary core area emerged in the suburban areas of Karachi (Figure 7).


**Table 6.** Urban landscape changes in Karachi during the periods of 1990–2000, 2000–2010, and 2010–2020.

**Figure 7.** Urban landscape in Karachi in (**a**) 1990, (**b**) 2000, (**c**) 2010, and (**d**) 2020.

The area of the suburban fringe and scatter settlements showed a marginal increase (Table 6). The urban open space increased from 121.8 sq.km in 1990 to 243.3 ssq.km in 2020, which indicates an increase in green space under the urbanized area. Most of this increase was observed in the core of districts of East, Korangi, and Kiamari. A drastic decrease from 3272 sq.km in 1990 to 2911 sq.km in 2020 in the area of non-urban open space can also be observed.

The changes in each urban spatial pattern class within districts were analyzed (Table 7). Between 1990 and 2020, the Kiamari, East, West, and Korangi districts had rapid growth in the primary urban core. From 1990 to 2020, no urban secondary core was found in the district of Kiamari, Central, South, West, or East, while this was observed in the districts of Malir and Korangi. The urban secondary core in the Korangi district merged with the urban primary core in 2020. The newly developed Malir district experienced high urban secondary core growth due to the large number of commercial and residential

−

−

− − −

developmental activities over the last two decades. The suburban fringe increased in the districts of Kiamari, West, East, and Malir, while the urban open space decreased within the Central and South districts. The decrease in open space in the CBD area might be attributed to the conversion of open space to residential and commercial lands.


**Table 7.** Urban landscape changes in different districts of Karachi (in percentage).

*4.3. Transition Probability Matrix Analysis*

The transition probability matrix was generated for the time periods of 1990–2000, 2000–2010, and 2010–2020 to demonstrate the probability that each land cover type was projected to change (Table 8). The values on the diagonal of the matrix represent the possibility of a land cover type maintaining its original state, and the values on the nondiagonal represent the possibility of a land cover type converting to other types. From 1990 to 2000, bare land was the most stable class with 0.77 probabilities, while the most dynamic class was cultivated land with transition probabilities of 0.32. From 2000 to 2010, water bodies were the most stable class with 0.67 probabilities, and grassland and shrub land were most dynamic with 0.20 probabilities. Similarly, from 2010 to 2020, mangroves were the most stable class with 0.76 probabilities, and cultivated land was the most dynamic class with 0.20 probabilities. The transition probability matrix from 2010 to 2020 was used to simulate the LULC map for Karachi city in 2030.


**Table 8.** Transition probability matrix of LULC classes in Karachi from 1990 to 2000, 2000 to 2010, and 2010 to 2020.

#### *4.4. LULC Simulation Results*

The validation results showed strong agreement with the simulation map (Table 9). The kappa values indicate that the CA–Markov model used is suitable for simulating future LULC maps in the study area.


**Table 9.** Validation results of the CA–Markov model.

The simulated LULC maps for Karachi city in 2030 are shown in Figure 8, and the changes for each LULC type are tabulated in Table 10. The simulation results show that the bare land area will significantly decrease in 2030 due to its conversion to a built-up area. The districts of Malir, South, and Kiamari are seeing the most growth in terms of urban built-up area. The spatial pattern of the predicted LULC indicated that the city's new residents would settle in sub-urban fringes surrounding the urban cores. Living in these areas allows them to be closer to work and facilitates a more convenient commute. Grassland and shrub land covered about 838.42 km<sup>2</sup> in 2020 and are expected to gradually increase to 999.06 km<sup>2</sup> in 2030.

**Figure 8.** LULC prediction results for Karachi city in 2030.


**Table 10.** Predicted LULC changes for Karachi city in 2030.

Although the validation results showed that the CA-Markov model was a reliable method for simulating land use change, there are several limitations in our study. Socioeconomic factors are among the most important variables influencing land use changes. Our study was unable to investigate several potential socioeconomic causes of urban expansion due to a lack of spatial data. Moreover, more sophisticated models can be developed to simulate urban growth in different areas of the study area [55]. Landsat images with a resolution of 30 m were used to construct the land-use/cover maps for LULC modeling. High-resolution satellite data may be employed in the future to generate more detailed observations of specific agricultural and urban covers.

#### **5. Conclusions**

An urban sprawl matrix methodology was used in this study to analyze changes in urban spatial patterns in Karachi over three decadal epochs (1990–2000, 2000–2010, and 2010–2020). The utilization of the urban sprawl matrix provided an accurate and effective assessment of Karachi's urban expansion tendencies. Future land cover changes in the study area were predicted using a CA–Markov model for 2030. The results indicate that the built-up area had expanded in a considerably unpredictable manner, which was mainly at the expense of agricultural land. The increase in mangroves and grassland and shrub land demonstrated the effectiveness of afforestation programs in improving vegetation coverage in the study area. Fast urban development was recorded in districts including Malir, West, and Kiamari from 1990 to 2020. The primary urban core expanded from the core districts, namely, the Central, South, and Eastern districts, and a new urban secondary core was observed in Malir in 2020. The LULC simulation results for 2030 revealed a significant increase in urban built-up area of 111.6% compared with that in 2020, mainly distributed in sub-urban fringes.

This study proved remote sensing and GIS techniques to be valuable tools in tracking and assessing changes in urban spatial patterns. The findings of the analysis can provide policy implications for future urban land transformation management and planning in order to achieve the Sustainable Development Goals. Future research could explore the forces that drive urban sprawl and examine how they interact with social, economic, and environmental repercussions in fast growing cities.

**Author Contributions:** Conceptualization, L.L. and F.C.; Methodology, M.F.B.; Software, M.F.B.; Validation, M.F.B.; Formal analysis, M.F.B.; Investigation, S.Q.; Resources, L.L.; Data curation, A.T.; Writing—original draft preparation, M.F.B.; Writing—review and editing, L.L., S.Q. and S.W.; Visualization, S.H.; Supervision, L.J.; Project administration, Q.L.; Funding acquisition, Q.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the National Key Research and Development Program of China (No. 2019YFD1100803).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Acknowledgments:** The authors are grateful for the comments from anonymous reviewers and the editors.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **Characterizing Urban Expansion Combining Concentric-Ring and Grid-Based Analysis for Latin American Cities**

**Su Wu <sup>1</sup> , Neema Simon Sumari <sup>2</sup> , Ting Dong <sup>3</sup> , Gang Xu 4,5,\* and Yanfang Liu <sup>1</sup>**


**Abstract:** Spatio-temporal characterization of urban expansion is the first step towards understanding how cities grow in space. We summarize two approaches used in urban expansion measurement, namely, concentric-ring analysis and grid-based analysis. Concentric-ring analysis divides urban areas into a series of rings, which is used to quantify the distance decay of urban elements from city centers. Grid-based analysis partitions a city into regular grids that are used to interpret local dynamics of urban growth. We combined these two approaches to characterize the urban expansion between 2000–2014 for five large Latin American cities (São Paulo, Brazil; Mexico City, Mexico; Buenos Aires, Argentina; Bogotá, Columbia; Santiago, Chile). Results show that the urban land (builtup area) density in concentric rings decreases from city centers to urban fringe, which can be well fitted by an inverse S curve. Parameters of fitting curves reflect disparities of urban extents and urban form among these five cities over time. Grid-based analysis presents the transformation of population from central to suburban areas, where new urban land mostly expands. In the global context, urban expansion in Latin America is far less rapid than countries or regions that are experiencing fast urbanization, such as Asia and Africa. Urban form of Latin American cities is particularly compact because of their rugged topographies with natural limitations.

**Keywords:** urban expansion; concentric-ring analysis; grid-based analysis; invers S curve; Latin America

#### **1. Introduction**

More and more people are now living in cities, driving the persistent expansion of urban land across the world [1]. Rapid expansion of urban land swallows cultivated land and natural land, threatening biodiversity and exacerbating environmental degradation [2–4]. Numerous studies have investigated spatio-temporal characteristics of urban land expansion and its environmental consequences [5,6]. Remote sensing and geographical information science (GIS) have greatly contributed to quantifying land use changes [7]. Remotely sensed imagery provides the first-hand data used to monitor urban growth in space [8]. GIS-based technologies provide a wealth of tools for urban expansion measurement [9].

The first step to measure urban expansion is its speed and intensity, for instance, the annual growth rate of built-up areas [10,11]. The intensity of urban expansion refers to proportions of urban land use changes to the total land area in a defined region [10,12]. Beyond the statistical description of urban land use changes, the spatial perspective is employed to answer the spatial pattern of urban form and where the urban expansion happens [13]. Thirdly, the temporal process cannot be separated when we characterize

**Citation:** Wu, S.; Sumari, N.S.; Dong, T.; Xu, G.; Liu, Y. Characterizing Urban Expansion Combining Concentric-Ring and Grid-Based Analysis for Latin American Cities. *Land* **2021**, *10*, 444. https://doi.org/ 10.3390/land10050444

Academic Editor: Iwona Cie´slak

Received: 17 March 2021 Accepted: 19 April 2021 Published: 22 April 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

urban expansion [14]. Landscape metrics are broadly used to measure dynamics and heterogeneities in the temporal process of urban expansion [15]. Finally, urban theories behind urban land use changes are proposed based on these quantitative measurements, such as the diffusion-coalescence theory [16–18], pattern-process interrelationships, distance-decay of densities [19,20], types of urban expansion (infilling, extension, and leapfrog) [21–24], and driving forces of urban expansion [25], etc.

As far as the spatial unit is concerned, there are two approaches to quantify urban expansion, namely, the concentric-ring analysis and grid-based analysis [13,26]. The concentric-ring analysis, also known as gradient analysis, divides urban areas into a series of concentric rings from the city center, which usually is the point of origin or central business district (CBD) of the city [27,28]. The concentric-ring analysis is often used in conjunction with landscape metrics (such as the percent of landscape, patch density, and many other indicators) [12,22]. They are used to analyze spatial patterns (distance-decay) of urban landscape. The distance-decay of urban land density (built-up density) in concentric rings reflects the gradient of urban development intensity [19]. Some studies divide concentric rings into multiple sectors to measure the heterogeneity of urban growth in different directions [15]. Considering irregular urban forms or spatial constrains, such as UK's 'green belts' and large water bodies in Wuhan, China, an improved partitioning method for concentric-ring analysis was proposed [12].

The grid-based analysis divides urban areas into regular grids; for example, 1 km grids [26,29–31]. Each grid is a sample and the urban dynamics in each grid are different, which determines the overall spatial pattern of land-use transformation. Grid-based analysis allows the correlation analysis between urban land and other occupied land, and it also builds the bridge between urban land expansion and population growth [26]. The growth in population fundamentally drives the expansion of urbanized land [32]. Numerous studies reported the faster growth rate of urban land than that of population over time, resulting in a decline in the urban population density [33–40].

In this study, we attempt to combine the concentric-ring analysis and grid-based analysis to quantify spatio-temporal characteristics of urban expansion. Our case study area is Latin America, defined as the Americas south of the United States. Latin America consists of 20 countries and 13 dependencies, with more than 640 million population in 2016 and an area of approximately 19,000,000 km<sup>2</sup> . In many countries of Latin America, the economy developed rapidly after the middle of the 20th century, and the level of urbanization has continued to increase. Latin America is one of the most urbanized continents in the world with almost 84% of the total population living in cities [41]. For example, Brazil's urbanization level is more than 85%, and Argentina's urbanization level is even more than 90% [42].

Although the urbanization level in Latin America is relatively high, compared with the extensive research of urban expansion in the United States and Europe, there are few studies on the temporal and spatial characteristics of urban expansion in Latin America [41]. In particular, there is a lack of comparative analysis of urban expansion and spatial dynamics among cities in Latin America [13,41,43]. Investigating the dynamics of urban land expansion in Latin America not only has regional significance, but also has international comparative value. It also provides precedents for developing countries that are undergoing rapid urbanization, such as Africa and Asia [44,45].

#### **2. Materials and Methods**

#### *2.1. Study Area*

New York University, United Nations-Habitat, and Lincoln Institute of Land Policy have published *The Atlas of Urban Expansion* (2016 Edition) of 200 cities around the world, which includes 26 cities in Latin America and the Caribbean [46,47]. Considering the representativeness of large cities and the evenness of their spatial distributions, we selected five large cities with a population of more than five million in 2014 from 26 cities, namely,

São Paulo (Brazil), Mexico City (Mexico), Buenos Aires (Argentina), Bogotá (Colombia), and Santiago (Chile) (Figure 1, Table 1).

**Figure 1.** Spatial distributions of five major cities in Latin America with their urban land use in 2000 and 2014.

**Table 1.** Urban population, built-up area, and population density in 2014 and their annual growth rates from 2000 to 2014 in five large Latin American cities.

