An explanation of urban sprawl phenomenon in Shiraz Metropolitan Area (SMA)

Cities xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Cities journal homepage: www.elsevier.com/locate/cities

An explanation of urban sprawl phenomenon in Shiraz Metropolitan Area (SMA) Bagher Bagheri, Sahar Nedae Tousi⁎ Department of Urban and Regional Planning, Faculty of Architecture and Urban Planning, Shahid Beheshti University, Tehran, Iran

A R T I C L E I N F O

A B S T R A C T

Keywords: Urban sprawl Per capita land consumption Path analysis Shiraz Metropolitan Area (SMA) in Iran

Urban Sprawl, as a low-density, unplanned, unlimited and sporadic physical expansion towards suburban area is one of the worldwide challenges facing spatial development planning in recent years. In a great part of literature on urban sprawl, dealing with this phenomenon depends on controlling two main factors of “population growth” and “per capita land consumption”. This study is to propose a comprehensive framework for dealing with this phenomenon emphasizing the case study of Shiraz Metropolitan Area (SMA) in Iran through identifying the drivers stimulating these two factors. Analyses were carried out by using spatial analytics, mathematical and statistical methods such as Holdern analysis, path analysis and other statistical analyses. Investigating the drivers and factors, the article suggests that unlike many reviewed experiences, “per capita land consumption” is not the main factor in SMA's Sprawl. Instead, “population growth” due to employment opportunities, higher relative household income and affordable housing policies are the main drivers. Furthermore, attracting creative class through development of knowledge economy and ICT infrastructures has adversely influenced urban sprawl. In addition, automobile-oriented developments have exacerbated this phenomenon by stimulating city expansion towards invaluable natural and rural areas. Thus, in order to control the phenomenon under study, it is necessary to take into account these factors in planning priorities and allocation of resources considering the causal relations between them.

1. Introduction Urban sprawl refers to low-density, poorly planned, auto-dependent and sporadic physical expansion of urban and rural area which spreads out over large amounts of rural land. In other words, it's the rapid expansion of residential and non-residential development to the relatively intact environment (Burchell & Galley, 2003; Ewing, 1997; Nelson & Duncan, 1995; USHUD, 1999). A review of the literature shows that this phenomenon is gobbling up of forests, farmland, wetlands and woodlands (Ermer, Mohrmann, & Sukopp, 1994; Leser & Huber-Frohli, 1997) and leads to the destruction of farmlands (Berry & Plaut, 1978; Fischel, 1982; Hasse & Lathrop, 2003b; Nelson, 1990; Zhang, Chen, Tan, & Sun, 2007), natural landscape, decreased desirability and viability (Akademie fur Raumforschung und Landesplanung (ARL) and Schweizerische Vereinigung fur Landesplanung (VLP), 1999; Landscape Gesellschaft fur GeoKommunikation, 2000–2002; Grimm, Grove, Pickett, & Redman, 2000), increased travel duration (Sierra Club, 1999), leads to soil, water and air pollutions (Jacquin, Misakova, & Gay, 2008; Stone, 2008; Wang, Zhu, Wang, & Shi, 2003; Weng, Liu, & Lu, 2007), increases energy and

⁎

natural resources consumption (Newman & Kenworthy, 1988) It's also increases costs of facility due to the spatial development of the city (Harvey & Clark, 1965), followed by increased current and civil costs and economic instability of the city and finally, decreased social ties and interactions leading to social instability (Benfield, Raimi, & Chen, 1999; Frumkin, 2002; Kunstler, 1993; Mitchell, 2001; Savitch, 2003; Sturm & Cohen, 2004; Yanos, 2007). Iran's cities have been faced with the urban sprawl phenomenon, especially since the 1970s. More recently, scientific studies have been proved negative impacts of urban sprawl in Iran's cities including the destruction of landscapes and natural resources around the city (Soltani, Hosseinpour, & Hajizadeh, 2017) and coastal areas and resulted in declining the declining tourism performance of the city (Dadras, ZulhaidiMohd Shafri, Ahmad, Pradhan, & Safarpour, 2014), degradation and destruction of agricultural land around the city, decline in productivity, and threat to food and economic security (Mohammadian Mosammam, Tavakoli Nia, Khani, Teymouri, & Kazemi, 2016), destruction and losing of groundwater resources (demolition of 88 flumes in Mashhad and 376 in Tehran (Hosseini et al., 2014) and the water crisis, pollution of water and soil,

Corresponding author. E-mail address: [email protected] (S.N. Tousi).

http://dx.doi.org/10.1016/j.cities.2017.10.011 Received 5 June 2017; Received in revised form 17 September 2017; Accepted 13 October 2017 0264-2751/ © 2017 Elsevier Ltd. All rights reserved.

Please cite this article as: Bagheri, B., Cities (2017), http://dx.doi.org/10.1016/j.cities.2017.10.011

Cities xxx (xxxx) xxx–xxx

B. Bagheri, S.N. Tousi

(Acioly & Davidson, 1996) low-density distribution of population, (Gouda et al., 2016) highly fragmented and disconnected households, shopping centers and workplaces (separate zoning of applications), (Akademie fur Raumforschung und Landesplanung (ARL) & Schweizerische Vereinigung fur Landesplanung (VLP), 1999) streets divided by large blocks with low-access of sidewalks (Al Gore, 1998) lack of a clear definition for thriving activity centers such as commercial centers and other urban centers. Burchell and Galley (2003), define urban sprawl as a sporadic low-density development determined by its indefinite spread out. In other words, urban sprawl is a significant residential and non-residential development in a relatively pristine environment. Jaeger et al. (2010), define this phenomenon as a visible status which is an urban landscape including sporadic urban development and building blocks placed in distance from one another. In addition to these quantitative definitions, there are some other qualitative definitions of urban sprawl. For instance, Al Gore (1998), presented one definition, by in his speech at the annual conference of the Democratic Leadership Council: “chaotic and poorly planned development that may make it impossible to greet neighbors on the sidewalk, it would require gasoline, a quarter of a gallon, to buy a bottle of milk, and it would be impossible for children to walk to their school”. Through an overview of these definitions, the study describes the typology of sprawl phenomenon in the following four possible systems: (Acioly & Davidson, 1996) spatial expansion system: chaotic and disordered growth, sporadic diffused growth (detached construction), distancing from the city center; (Gouda et al., 2016) urban planning system (level of commitment to the plan's principles): unplanned and uncontrolled growth, dispersion of development beyond urban and functional edges, (Akademie fur Raumforschung und Landesplanung (ARL) & Schweizerische Vereinigung fur Landesplanung (VLP), 1999) the system of land-use and density: use change of open spaces, farmlands and rural fields, single functional development and promoting zoning approach, low density development; (Al Gore, 1998) system of communication and accessibility: limited and invariant access to transportation (automobile-oriented development), distancing residences from workplaces and activity centers. Aforementioned definitions are classified in Table 1 based on these four selected systems.

increasing the cost of providing urban services, increasing the time and length of intra-city trips (Mohammady & Delavar, 2014) and as a result, increasing the consumption of fuel and energy as well as changing the local climate (Zanganeh Shahraki et al., 2011), Social segregation (Eslami Mahmoudabadi, Soroushnia, & Zekri, 2013; Zali, Hashemzadeh Ghal'ejough, & Esmailzadeh, 2016) and the reduction of social capital and increasing the crime rate of the city; There are two major theoretical approaches to urban sprawl which are both opposite and supportive: anti-sprawl and pro-sprawl. Antisprawl movement is classified in two categories of “protection” and “smart growth” including three branches: advocates of more efficient urban planning in terms of energy, high-density supporters and supporters of earth, valuable rural areas and environmental assets. The aims of these branches have led to define and measure the sprawl from different point of view. For example, high-density supporters define and measure urban sprawl phenomenon in terms of density rate in metropolitan areas. “USA today's sprawl index” is an example in this regard. However, environmentalists, measure urban sprawl via agricultural and horticultural land-use change outside the metropolitan area (Beck, Kolankiewicz, & Camarota, 2003; Nelson, 1990). On the contrary, there is a group of liberal politicians, researchers, and journalists that have promoted urban sprawl phenomenon and American customs and values. They believe that urban sprawl has many benefits such as affordable housing, free parking lot, free movement, enough space, yard and neighborhood with green areas around the city as well as high quality of life for citizens who are tired of city life (Beck et al., 2003; Kahn, 2001). Similar to other studies in Iran, this study also uses “Anti-sprawlers” approaches and aims to introduce comprehensive framework deals with this phenomenon by conducting case study of Shiraz Metropolitan Area (SMA) in Iran. Accordingly, with the aim of presenting a general framework for Iranian metropolises, an attempt has been made to make an analogy between the patterns and the reasons of sprawl in other cities of Iran, including large, medium and small cities based on existing studies and researches. Shiraz as the sixth largest metropolis areas in Iran, has been chosen to explain urban sprawl because of its dramatic changes during past five decades. In this regard, after conducting the spatial analytics of urban sprawl in SMA, the causal relation explaining this phenomenon would be explored using path analysis method. This method is used to describe the direct and indirect dependencies among a set of variables or drivers stimulating the sprawl.

2.2. Urban sprawl measurement Different methods have been introduced in the literature such as single-measure or multi-measure methods for measuring urban sprawl. However, the simplest and the most frequent variable is density (Peiser, 1989). In addition to density, other variables are also employed in urban sprawl measurement, such as location of residential and activity clusters in relation to each other or to other centers, continuity rate of urban developments, centralization and mixed-use (Bertaud & Malpezzi, 1999; Ewing et al., 2002; Galster et al., 2001; Gordon & Richardson, 1997). Kahn (2001), using “distribution of job centers in relation to city center” as his basic measure, has investigated the concentration of all job centers within 10 miles around City Business District (CBD) as zero sprawl and all job centers outside this radius as full sprawl. However, his definition is practically indifferent to the status of residential centers and porosity patterns of development for activity centers. Also, Glaeser, Kahn, and Chu (2001), have used the same measure to examine the sprawl of activity centers in USA in three miles and ten miles from businesses center. Many researchers (Batty, Xie, & Sun, 1999; Siedentop & Fina, 2012; Sudhira, Ramachandra, & Jagadish, 2004; Terzi & Bolen, 2011; Torrens & Alberti, 2000; Tsai, 2005) used these three factors in urban sprawl measurement (Acioly & Davidson, 1996) density (gross population and construction density and occupancy level), (Gouda et al., 2016) mixed-use ratio (residential use ratio, commercial use ratio, urban facilities ratio such as hospitals, schools and parks) and (Akademie fur Raumforschung und Landesplanung (ARL) & Schweizerische Vereinigung fur Landesplanung (VLP), 1999)

2. Theoretical framework 2.1. Conceptualization of urban sprawl In contrast to “compacts cities”, “urban sprawl” is an ambiguous concept and researches and organizations have not yet developed a commonly accepted definition of what constitutes urban sprawl. Urban sprawl could be defined as a situation (by measuring the degree of sprawl) or a process. In general, some experts claim urban sprawl is an urban/rural spatial expansion that constitutes three major features: (Acioly & Davidson, 1996) sporadic or dispersing development (Gouda, Hosseini, & Masoumi, 2016) commercial strips development (Akademie fur Raumforschung und Landesplanung (ARL) & Schweizerische Vereinigung fur Landesplanung (VLP), 1999) wide spread of low-density or single functional development (Ewing, 1997; Hasse & Lathrop, 2003a; Jaeger, Bertiller, Schwick, Cavens, & Kienast, 2010; Sierra Club, 1999). Also, according to Ewing, Pendall, and Chen (2002), this expression is recognizable by indicators such as limited-access to facilities and services and lack of functional outdoor spaces. Yet from another perspective, urban sprawl is the process by which the dispersion of development across the region happens faster than population growth (Beck et al., 2003; Ewing et al., 2002; Fulton, Pendall, Nguyen, & Harrison, 2001). Therefore, it has four dimensions: 2

Cities xxx (xxxx) xxx–xxx

B. Bagheri, S.N. Tousi

Table 1 Theoretical categorization of reviewed definitions for urban sprawl phenomenon. Categories

Characteristic

Sources

Spatial development system

Chaotic and disordered growth

Uneven growth (Nelson A., 1999; Akademie fur Raumforschung und Landesplanung, 1970), Chaotic development (Al Gore, 1998) Sporadic density (Akademie fur Raumforschung und Landesplanung, 1970; Ewing, 1997; USHUD, 1999), Excessive and dispersed construction (Jaeger et al., 2010), Low value of clustering and continuity (Galster et al., 2001; Nelson, 1999) Leap away from the center (Burchell & Galley, 2003), low centeredness (Galster et al., 2001), Distribution of employment more than ten miles from the CBD (Kahn, 2001), Distribution of employment beyond a three-mile radius from the CBD (Glaeser et al., 2001) Non-systematic housing locations (Akademie fur Raumforschung und Landesplanung, 1970), Nonsystematic expansion (Ermer et al., 1994), Intensified unplanned and unsystematic growth outside (Leser & Huber-Frohli, 1997; Geo-Kommunikation, 2000–2002), Destructive plans for surrounding areas (Al Gore, 1998; Nelson, 1999), Poor planning (Moe, 1999) Overflow of residential areas (Ermer et al., 1994), Extending out the edge of service and employment (Leser & Huber-Frohli, 1997; Sierra Club, 1999), Unlimited expansion outside (USHUD, 1999; Akademie fur Raumforschung und Landesplanung (ARL) & Schweizerische Vereinigung fur Landesplanung (VLP), 1999; Geo-Kommunikation, 2000–2002) Land consumption being faster than population growth (Ewing et al., 2002; Fulton et al., 2001), Increased per capita land consumption (Beck et al., 2003) Excessive use of open landscapes (Akademie fur Raumforschung und Landesplanung, 1970; Ermer et al., 1994), Destruction of natural landscapes and ecosystems (Akademie fur Raumforschung und Landesplanung (ARL) & Schweizerische Vereinigung fur Landesplanung (VLP), 1999), Leading urbanization towards rural areas (Geo-Kommunikation, 2000–2002) Separation of the places for residence, shopping, work, leisure and education (Akademie fur Raumforschung und Landesplanung, 1970; Ewing, 1997; Sierra Club, 1999), Differentiation of landuses (USHUD, 1999), Parts being non-performative (Geo-Kommunikation, 2000–2002), Mixing and proximity being at a low degree (Galster et al., 2001) Low-density zoning (Burchell & Galley, 2003; Ermer et al., 1994; Ewing, 1997; Ewing et al., 2002; Galster et al., 2001; Pendall, 1999; USHUD, 1999) Excessive use of private cars (Ewing, 1997; Sierra Club, 1999), Difficult access (Al Gore, 1998; Ewing et al., 2002; USHUD, 1999), Automobile-oriented development (Moe, 1999) (Ewing et al., 2002; Galster et al., 2001; Sierra Club, 1999)

Sporadic diffused growth (detached construction) Distancing from the city center

Planning system

Unplanned and uncontrolled growth

Dispersion of suburban development

System of land-use and density

Physical growth surpassing population growth Use change of outdoors, farmlands and rural fields

Single functional development

Low density System of communication and accessibility

Limited and invariant accessibility Distancing residences from workplaces

Mohammady & Delavar, 2014; Mohammadian Mosammam et al., 2016), investigating the relationship between climatic criteria and sprawl using methods such as Pearson correlation and multiple regression (Roshan, Zanganeh Shahraki, Sauri, & Borna, 2010), using the Particle Swarm Optimization algorithm (Mohammady & Delavar, 2014) are among the most common methods can be mentioned. In general, using this method, one may achieve different results about sprawl through the comparison of urban and non-urban areas during different periods of time. Among various measuring methods for sprawl in metropolitan areas, measuring per capita land consumption distancing from CBD, land cover changes, the amount and spatial distribution of population are the most common ones which have been used in the present study in order to assess the sprawl. But it is noteworthy that this paper is not focused on different measuring methods, in fact, this study is to prove the existence of urban sprawl, and also, the most important purpose of this paper is to discover the casual relations between the drivers of sprawl.

compactness rate of urban areas which is measured by distance (average geographical distance between neighborhoods, average distance between neighborhoods and centers, average distance between job centers and CBD, etc.). In many studies conducted on different spatial dimensions, different mathematical methods were used to measure the spread of factors and complications, including Average Nearest Neighbor (ANN), Shannon's entropy, central feature, Spatial autocorrelation (Moran's I), cluster and outlier analysis (Anselin Local Moran's I), Geary, high/low clustering (Getis-Ord General G), hot and cold spot analysis, which have been used in several preceding texts (Carnes & Ogneva-Himmelberger, 2012; Conley, Stein, & Davis, 2014; Hajizadeh, Campbell, & Sarma, 2015; Porat, Shoshany, & Frenkel, 2012; Shariat-Mohaymany & Shahri, 2016; Torrens, 2008). In another study, Yeh and Li (2001) have measured sprawl through the analysis of land cover. This type of study, which is one of the most common measuring methods for sprawl, has been also used in some other studies (Feng, Du, Zhu, Luo, & Adaku, 2016; Jat, Garg, & Khare, 2008; Tewolde & Cabral, 2011). In order to measure this phenomenon different methods were used in different studies of sprawl in Iran. Descriptive and qualitative studies (such as asking the experts of the government agencies such as the municipality, the water and wastewater organization, and the transportation sector) and field observations (Masoumi, 2012b; Mohammadian Mosammam et al., 2016; Soltani et al., 2017), investigating aerial and satellite images in different periods (Asgarian, Jabbarian Amir, Alizadeh Shabani, & Feghhi, 2014; Dadras et al., 2014; Mohammadian Mosammam et al., 2016; Mohammady & Delavar, 2014; Zanganeh Shahraki et al., 2011), Holdren (Mohammadi, Zarabi, & Mobaraki, 2012; Soltani et al., 2017; Zali et al., 2016), combining the multi-criteria evaluation method and Cellular automataMarkov model (Asgarian et al., 2014), Shannon's entropy (Asgarian et al., 2014; Soltani et al., 2017; Mohammadi et al., 2012;

2.3. Urban sprawl causes (theoretical roots) A review of literature to explain urban sprawl phenomenon gives a number of drivers. Two main factors, however, may be deduced from a general perspective (Beck et al., 2003): first, “population increase” due to natural growth and immigration; second, “per capita land consumption” or “per capita sprawl” (urban sprawl rate as a function of population multiplied by per capita land consumption). In addition to variables of population and per capita land consumption, an overview of theoretical discourses on urban sprawl will provide us with other drivers (Ewing, 1997; Bhatta, 2010; Geo-Kommunikation, 2000–2002; Harvey & Clark, 1965; Cheng & Masser, 2003; Yang & Lo, 2003). These include seven main factors related to the following concepts (Acioly & Davidson, 1996) economics of land and housing, (Gouda et al., 2016) land-use and lack of available spaces within urban context, (Akademie fur 3

Per capita land consumption

Population Growth

Endogenous variables

4 Technological expansions

Technological advancements in information and communication Economic structure of cities

Resulting in poor physical media such as rugged lands, wetlands, mineral lands, water lands etc.

Automobile-oriented-ness

Natural geographical factors

(Bhatta, 2010; Moe, 1999)

Knowledge Based Development Large-scaled Industrial Units

IT Development

Municipal Fund Social Housing Policies

Transportation preferences

(Slack, 2002) (Aurand, 2013; Barnes et al., 2001; Bhatta, 2010) (Johnson, 1991; McGahey, Malloy, Kazanas, & Jacobs, 1990) (Yigitcanlar, 2007) (Bhatta, 2010; Leser & Huber-Frohli, 1997)

Creative class

(Flew, 2012; Florida, 2002a, 2002b; Scott, 2006) (Bhatta, 2010)

Governmental performance

Non-food costs/Housing costs

Rate of urbanization

Single-family home Per capita land consumption/The number of real estate transactions Population Growth Housing Pattern

The price of land and housing Expansion of street networks

Household income

Speculations of land and housing

Indicators

(Bhatta, 2010)

(Beck et al., 2003; Bhatta, 2010) (Acioly & Davidson, 1996; Barnes et al., 2001; Bhatta, 2010) (Bhatta B., 2010; Bekele, 2005)

(Bhatta, 2010; Clawson, 1962; Harvey & Clark, 1965) (Bhatta, 2009; Boyce, 1963; Giuliano, 1989; Oueslati, Alvanides, & Garrod, 2015) (Bhatta, 2010; Brueckner & Kim, 2003) (Cheng & Masser, 2003; Harvey & Clark, 1965; Yang & Lo, 2003) (Acioly & Davidson, 1996) (Beck et al., 2003; Harvey & Clark, 1965)

Sources

Centrifugal forces

Knowledge economy Industrialization

Government developmental policies Local budgets Affordable housing

Population Growth Demand of more living space Tendencies for urbanization Household Consumption Patterns Creative class absorption

Low-rise development Land consumption

Housing Investment Network Development

Purchasing Power

Speculations

Drivers

Urban development system of planning and policymaking

Cultural trends and behavioral patterns

Land-use planning and network development

Economics of land and housing

Categories

Exogenous variables

Table 2 Conceptual framework of urban sprawl causal model in metropolitan areas.

Lack of/poor planning and control mechanisms of development guidelines in cities Provision of land and housing by governmental and local bodies Tax policies for different urban areas Development of communication networks (transportation) Development of IT infrastructures Change in economic structures of cities towards industries and land consuming activities Expansion of information and communication technologies as well as restructuring the economy towards knowledge-based activities Increased negative externalities in urban economies due to the increased population and activity, high density of population, traffic, pollution and congestion

Increased tendency towards living in nature to avoid urbanization Natural increase in population and immigration Increased tendency towards single-family patterns of constructions Increased tendency towards land and housing ownership/ investments in land and housing Increased use of private cars

The absence or shortage of open spaces for development and difficulty of land acquisition Pro-sprawl land use planning and network development pattern

High price of land and housing in the inner areas and low price in urban fringes Legal property disputes on urban lands Housing economy

Rationale

B. Bagheri, S.N. Tousi

Cities xxx (xxxx) xxx–xxx

Cities xxx (xxxx) xxx–xxx

B. Bagheri, S.N. Tousi

during the last five decades. In addition to the two points are shown in Fig. 2, many other regions of the study area such as settlements, university, shopping mall and even a new city have been developing this way over time. Therefore, in this area, in addition to the rapid growth of settlements, the creation of new points is also a clear reason of the sprawl and its various species. Fig. 3 shows the different periods of the establishing of city. Following this introduction, SMA's urban sprawl would be analyzed and modelled based on available data in Iran.

Raumforschung und Landesplanung (ARL) & Schweizerische Vereinigung fur Landesplanung (VLP), 1999) cultural trends as well as consumption and behavioral patterns (Al Gore, 1998) urban development system of planning and policymaking, (Anselin, 1995) advancements in information and communication technology leading to spatial agglomeration of population and activities being less important, (Asgarian et al., 2014) changes in the economic structure of the cities and (Aurand, 2013) development of centrifugal forces and negative externalities in urban economies due to the vast scale of large cities. Although natural geographical factors are important causes of sprawl, they would not be a matter of consideration in this article. Although, according to the article's conceptual framework is shown in Table 2, these factors are considered as exogenous variables leading to “population and per capita land consumption”, as two main factors or endogenous variables, which eventually cause urban sprawl. Based on the proposed conceptual framework, tracing the relation between endogenous and exogenous variables of sprawl, researchers are to propose a theoretical plausible causal model. Through the steps ahead, this model would be tested using path analysis methods aimed to find the best fitted model according to the empirical data of SMA's sprawl. For this purpose, the following steps would be taken: (Acioly & Davidson, 1996) the typology and status of urban sprawl as well as its changes in SMA, Iran, from 1956 to 2013 would be discussed (Gouda et al., 2016). Following that, the article would present theoretical roots along with an explanation of urban sprawl phenomenon in SMA in order to provide instructions and tips for dealing with the problem through identifying and indexing its drivers (Akademie fur Raumforschung und Landesplanung (ARL) & Schweizerische Vereinigung fur Landesplanung (VLP), 1999). Afterwards, path analysis would be the selected method to test the conceptual model derived from a review of related literature and also, to suggest an experimental good-fitted model for SMA (Al Gore, 1998). Eventually, based on the best fitted model explaining urban sprawl in SMA, effective solutions for dealing with sprawl would be discussed.

4. Methods Methods are used in this paper include three distinct categories. The first issue is the methods to prove the existence of and measuring the phenomenon of sprawl, the second issue is the methods to identify the spatial principles of sprawl, finally, the third issue is the methods to explain the phenomenon of sprawl with the intention of roots and identify drivers that influencing its occurrence. The following describes the methods used in the paper. 4.1. The methods used to prove the existence of and measuring the sprawl in Shiraz Metropolitan Area (SMA) This paper aims to prove the existence of sprawl, the first step is to use the aerial imagery analysis related to different periods of time. Using this method, changes in land cover during different years will be monitored. In addition to the spatial analyses, mathematical methods are used to quantify and measure the intensity of the sprawl. Therefore, the coefficients of Average Nearest Neighbor (ANN) and Spatial autocorrelation (Moran's I) in GIS software and Shannon's entropy are also calculated and analyzed in Excel software to show different dispersion situations in SMA from 1956 to 2011. The purpose of the application of several mathematical methods is to prove the existence of sprawl overhead in both physical (ANN) and demographic (Moran's I). In this regard, the most commonly used method was used (Shannon's entropy) which is accepted in most studies. Thus, “Average Nearest Neighbor (ANN)”, “Spatial Autocorrelation (Moran's I)” coefficients and “Shannon's entropy” will be used in order to display the status of urban sprawl in SMA and its changes from 1956 to 2011. These coefficients are described as follows: Average Nearest Neighbor:

3. An introduction to selected case study: the status of urban sprawl in SMA Shiraz, as capital of Fars Province, is one of the major poles of development in Iran. Shiraz also is one of the metropolises in Iran and one of the most important tourism cities in the region of South Zagros. Its population totaled more than 1, 781, 707 people in 2016. The metropolitan area in the foregoing research is a region consisting of a major populated urban core, with urban and rural settlements, and an activity surrounding it that has a high level of functional and physical attachment to the central core of the city. In fact, this concept is based on the scope of the labor market, which in some way is defined as the main focus of activity and employment (a region with a lot of available jobs) and its adjacent areas, which have strong links with the center. (See Fig. 1.) As the sixth metropolis, this area has been chosen to investigate and explain urban sprawl because of its dramatic changes during the past five decades. These changes were due to its twenty times increased area towards the surrounding horticultural lands. The expansion of the city has increased very fast since 1950s, so that it has covered some rich and valuable surrounding horticultural land in both connected and disconnected ways in different periods. This happened due to the topographical conditions, especially along the suburban main roads leading to the city. A review of aerial photos of the area in recent years is represented in Fig. 2. This figure represents two points that have changed a lot. In this figure, it can be clearly seen the rate of conversion of non-urban land to urban and the growth rate of settlements over the past few years. Furthermore, these reviews highlight the expansion of the area from 894 ha in 1956 to about 18,000 ha in 2011 (more than twenty times increase). Thus, it suggests the intensified urban sprawl phenomenon

∑in= 1 di n 0.5

ANN =

n A

(1)

where, di = distance of feature i from the nearest feature, n = number of features, A = area of the region under study. If ANN coefficient is less than 1, the spatial development pattern is clustering; otherwise it is dispersed spatial development pattern (ESRI, 2009). Spatial Autocorrelation (Moran's I): n

I=

n

n ∑i = 1 ∑ j = 1 wij (yi − y )(yj − y ) n

n

n

(∑i = 1 ∑ j = 1 wij )(∑i = 1 (yi − y )2)

(2)

where, n = number of sub-areas, yi = population or employment in sub-area i, yj = population or employment in sub-area j, y is the mean of population or employment, wij = weighting between two sub-areas i and j. Moran coefficient value is varying from −1 to +1 and its high value means that spatial development pattern is clustering. While, if it is zero it gives a random pattern, and its low value gives a dispersed spatial development pattern (Anselin, 1995). Shannon's entropy: n

Hn =

∑ Pi i

5

1 log ⎛ ⎞ ⎝ Pi ⎠

(3)

Cities xxx (xxxx) xxx–xxx

B. Bagheri, S.N. Tousi

Fig. 1. Fars province in Iran (left), SMA (right).

time-series analysis of the relation between the two indicators: “the distance of residential and activity centers from the city center” and “the proportion of population changes to the changes of the area”.

where, Pi is the proportion of the variable in the ith zone (proportion of the built up area in each zone) and n is the total number of zones (Yeh & Li, 2001). We have used relative entropy for scale the entropy value from 0 to 1. The relative entropy (H'n) is (Thomas, 1981): n

H′n =

i

4.3. The methods used to explain the sprawl and the contribution of different drivers in Shiraz Metropolitan Area

1

∑ Pi log ⎛ Pi ⎞/log(n) ⎝

⎠

(4) There are different approaches to explain urban sprawl and, also, to identify and quantify different components or exogenous variables listed in Table 2, affecting urban sprawl phenomenon:

Shannon entropy is used to measure the compression and distribution of a geographic variable (urbanized lands) (Theil, 1967; Thomas, 1981); if the distribution is concentrated only in one zone, it will be zero; and as the amount of the distribution gets closer to one, the distribution is considered more balanced (Yeh & Li, 2001).

1.1.1. First approach-Simple Ratio Approach, which calculates simple sum of components' growth rates categorized in Table 2 as exogenous variables. The proportion of each variable's growth rate to total growth rate gives the rate of impact on dependent or endogenous phenomenon (urban sprawl).

4.2. The methods used in the study of spatial patterns of sprawl In addition to the above-mentioned mathematical methods, the spatial patterns of sprawl in this area have been studied through the

Fig. 2. Aerial images of two main points of region, left images related to left point in the map and right images related to right point in the map.

6

Cities xxx (xxxx) xxx–xxx

B. Bagheri, S.N. Tousi

Fig. 3. Urban growth and expansion of Shiraz (1956–2011).

drivers are not always independent of each other; they influence each other and are influenced by other drivers as well. There are several multi-collinearity methods to explain the effects of sprawl's drivers among which are path analysis, factor analysis, principal component analysis, etc. that have been used in various studies (Cervero, 2003; Guhathakurta & Gober, 2010; Inostroza & Helena Tábbita, 2016; Kaur & Singh, 2005; Riitters et al., 1995; Sanders, Zuidgeest, & Geurs, 2015; Straus, Chang, & Hong, 2016). Thus, through the above-mentioned methods, path analysis makes it possible to discover causal relations in the path form. It also makes it possible to explain the phenomenon based on multiple variables other than population and per capita land consumption. Using the methods mentioned above by SPSS software, this study attempts to explain urban sprawl phenomenon in SMA through multiple variables other than “population” and “per capita land consumption” by taking the following successive steps:

1.1.2. Second approach-Holdren method: John Holdren presented this method to calculate the proportion of an independent variable (population and per capita land consumption) influencing the changes of dependent variable (sprawl). In this method, exponential assumption is used through following formula, rather than simple linear assumption (Holdren, 1991):

ln (land− use in current year) ln (land− use in the base year) ln (per capita land consumption current year ) = ln (per capita land consumption in the base year) ln (final population ) + ln ( base population ) In this method, based on the literature reviewed, the underlying assumption is that the two variables of “population growth” and “per capita land consumption” are the major factors involved in the occurrence of sprawl. In this paper, in addition to this method, multiple linear regressions between population growth, per capita land consumption and sprawl has also been used in order to identify the impact of each independent variable on the dependent variable of sprawl in Shiraz Metropolitan Area. Despite the popularity of this method in the literature of sprawl, based on the conceptual framework outlined in Table 2, the authors believe that other exogenous variables or factors are indirect factors affecting the sprawl phenomenon that should be recognized. Accordingly, the third approach has also been used as below.

a) First step: Measuring urban sprawl in SMA and it changes in time series of 1956 to 2012 by spatial analytics and mathematical methods mentioned above; b) Second step: Indexing drivers affecting urban sprawl phenomenon in SMA based on the conceptual framework summarized in Table 2 and identification of theoretical causal relations by defining input, output and outcome variables. c) Third step: proposing a determining theoretical model of urban sprawl phenomenon in the region; d) Fourth step: to test the theoretical model and develop most fitted empirical model of urban sprawl phenomenon in metropolitan area of Shiraz, using path analysis.

1.1.3. Third approach-explorative research (Path Analysis): In fact, known

The following outlines the issues to be addressed. (See Fig. 4.)

7

Cities xxx (xxxx) xxx–xxx

B. Bagheri, S.N. Tousi

Fig. 4. The methodological flowchart of this paper.

5. Analyses and results

II, the routes became almost semi-gridiron, and after the 1980s they became complete gridiron, which resulted in the leapfrogged development in a small-scale, and after this stage the connectivity and adhesion of the village, especially the old villages to the city happened. Unlike western patterns, the sprawl patterns resulting from the commercial strip development also rarely seen in Iranian cities. In addition to the mentioned cases, this paper also shows evidence of transnational and inter-regional differences patterns in large, medium and small cities in Iran. Although a review of existing studies in this area shows that most of the attention is focused on the horizontal unplanned growth of large cities, but middle and small cities are also faced with this problem. According to most studies, the sprawl of large cities over 500,000 population in Iran has been largely influenced by the common patterns of management and implementation of urban development projects, as well as the out posting of the immigration rates because of attractiveness places as compared to the urban supply rate, as well as the provision of suitable housing centers follows the horizontal,fragmented, disproportional, scattered development pattern and leapfrogged resulting from the adhesion of the village to the city mainly along the

The comparison of the results of the sprawl pattern study in Shiraz Metropolitan Area with other western samples and internal samples reveals differences and similarities. Masoumi (2012b), illustrates a number of major differences between the sprawl patterns of Iran and Western countries: the urban sprawl in North America and Western Europe have a large link with the planned development of urban suburbs that the term “suburban sprawl” is used. However, the shape of rapid change and sprawl developments around Iranian cities is not related to the planned development of the suburbs; on the other hand, single-use development under the influence of zoning laws in the West is also one of the first factors leading to sprawl in this regard; But since development plans in Iran have not been influenced by zoning laws, so this factor cannot be cited as one of the most important factors leading to sprawl. Another difference is in the type of street network in the cities of Iran and the West. In Iran, especially in historic cities such as Yazd and Kashan, the paths were initially stalled and curved, and then these routes became easiest for the car after the 1930s. After World War 8

Cities xxx (xxxx) xxx–xxx

B. Bagheri, S.N. Tousi

Fig. 5. Urban sprawl typology of activity and residential centers around Shiraz.

rural source had been as a dominate pattern in large cities of Iran, but the sprawl pattern of small and medium-sized cities such as Kashan (Masoumi, 2012a), Kazeroun (Soltani et al., 2017) and Sirjan (Eslami Mahmoudabadi et al., 2013) is largely influenced by planned developments as a result of government decision making in the form of new unestablished centers with lacking infrastructure and neighborhood centers. Given that the agenda on this paper is to prove the existence of and measure the sprawl in SMA and also, to identify and assess the effects of sprawl's drivers thus, after presenting the results and findings derived from the spatial typology and measuring the sprawl, we will identify and measure the causal relations between the drivers.

main road networks around the city. Zali et al. (2016), studied the dominant pattern of sprawl of Tehran metropolitan, Hosseini et al. (2014), in relation to the pattern of sprawl of Mashhad with integration of approximately 9 rural to urban areas between 1966 and 1976 and the study of Dadras et al. (2014), which was conducted in Bandar Abbas and it was affected by natural limitations and structure as military zones, rocky, coastal and natural slope of land, as well as connecting rural areas such as Soru, Shahrak Shagho and Nakhl-e-Nakhoda and also the findings of Zanganeh Shahraki et al. (2011), in Yazd were included in this category. In the case of some other large cities such as Qom, the sprawl caused by the effects of the proximity to the two metropolises of Tehran and Isfahan, as well as the lack of fertile lands and water resources on the periphery of the city and its hinterland has led to a pattern of linear development along the outskirts of the city (Mohammadian Mosammam et al., 2016). The sprawl pattern with the 9

Cities xxx (xxxx) xxx–xxx

B. Bagheri, S.N. Tousi

Fig. 6. Urban sprawl phenomenon and its changes in SMA from 1956 to 2011.

Table 3 Changes in urban sprawl of Shiraz Metropolitan Area from 1956 to 2011 using two ANN and Moran coefficients.

Year Desc. Urban sprawl analysis

1956

1976

1996

2011

Moran's Index

0.016870

0.011648

0.007189

0.006078

0.374824 0.546851168

0.411527 0.343401689

0.508543 0.160900606

0.601727 0.235996607

Urban sprawl analysis

ANN Index Shannon's entropy (H')

5.1. Proving the existence and measuring the phenomenon of sprawl in Shiraz Metropolitan Area (SMA)

Raumforschung und Landesplanung (ARL) & Schweizerische Vereinigung fur Landesplanung (VLP), 1999) isolated and dispersed leapfrog expansion of activity and residential centers around this metropolis and (Al Gore, 1998) continuous expansion or extension (Fig. 5). In addition to the above, Fig. 5 clearly shows the role of geographical and topographical factors around the study area in determining the incidence of sprawl. As Harvey and Clark (1965), and

Generally, four types of sprawl are identifiable in the Shiraz Metropolitan Area through subjective assessments (Acioly & Davidson, 1996) linear: emphasizing North-West territories, (Gouda et al., 2016) clustering: emphasizing residential town of Sadra, (Akademie fur 10

Cities xxx (xxxx) xxx–xxx

B. Bagheri, S.N. Tousi

Fig. 7. The relation between variables of “distance of residential and activity centers from the city center” and “change of population compared to the change of area from 1956 to 2011”.

in Fig. 8, urban sprawl has increased more quickly in Shiraz Metropolitan Area during the time periods of higher population growth rate. Many, including urban development planners, believe that population growth does not play a role in urban sprawl. The reason for this claim is that, sometimes urban sprawl is seen in areas with stable populations and in many cases, urban sprawl rate is higher than population growth rate. Also, there is a lack of reference to the drivers of population growth in different texts. One of the main reasons for this is the inability of public policy to control this phenomenon (Beck et al., 2003). as shown in Fig. 9, time series of “population” and “per capita land consumption” impacts on Shiraz Metropolitan Area's sprawl revealed a gradual increase of “per capita land consumption” impact, during these years, in contrast to relative lower its size than population. Following, the effects of other variables in addition to these two variables, as the leading research agenda for this study, would be discussed. Also, these latter variables in theoretical foundations of urban sprawl phenomenon have identified in Shiraz Metropolitan Area. As suggested before, urban sprawl phenomenon in SMA should be explained through multiple variables other than population and per capita land consumption by taking the following successive steps:

Barnes, Morgan, Roberge, and Lowe (2001), acknowledged the role and effect of natural and geographical factors on sprawl, this issue is clearly evident in the study area of this study because sprawl towards agricultural land and plains, especially in the southern region of the study area, is more severe than other areas. In Fig. 6, one can clearly sense the growth of urban residents and, in other words, the conversion rate of non-urban areas to urban areas between 1956 and 2011. Based on this figure, the growth of these urban areas is more tangible in North West than in other sectors. Table 3 shows these changes through mathematical calculations. In addition to the spatial analysis verifying the occurrence of different types of sprawl in Shiraz Metropolitan Area and the exacerbation of the phenomenon over the time, mathematical analysis through Moran indices, ANN and Shannon's entropy as well as the results of Table 3 confirms sprawl in this area. Although the Moran's Index under investigation in each time period always shows a random pattern, changes made from 0.0168 in 1956 to 0.006 in 2011 represents a sprawl pattern. This claim may be proved again via ANN coefficient which has increased from 0.37 in 1956 to 0.60 in 2011. The decreasing value of Shannon's entropy index from 1956 to 1996 shows that the first has a more appropriate spatial distribution than the latter one; as closer we get to 1996, the city of Shiraz gets more focused and dominant and therefore the spatial balance is more challenged (reduced value of Shannon's entropy index). With the advent of the new city of Sadra in 1996 and its growth until 2011, the dominant power of Shiraz gets reduced (increased value of Shannon index). However, Shiraz continues to rule the region as an absolute power. Comparative studies of urban sprawl changes in residential and activity clusters around SMA (Fig. 6 and Table 3) from 1956 to 2011, and the analysis of its status based on the variable of distance from the geometric center represented in Fig.7 indicate that, in total, settlements located near the metropolis have faced more sprawl over these years. "Average annual growth rate of area in relation to annual population growth rate" is considered as an indicator for this investigation.

1.1.4. First step: urban sprawl measurement in SMA in time series of 1956 to 2012; as previously described, the study identifies urban sprawl through some features. In this step, relevant data would be collected and reconstructed in time series before selecting indicators of urban sprawl phenomenon according to the data provided. Interpolation method has been used for reconstruction due to the limited data time series provided by General Population and Housing Censuses from 1956 to 2012 and also the need to draw relevant data as time series. 1.1.5. Second step: identifying and indexing drivers affecting urban sprawl phenomenon in SMA; again, in this step, after selecting indicators representing the effective drivers on urban sprawl in SMA (Table 5), the study has interpolated and extrapolated the relevant data using nonlinear regression analysis, due to the lack of time series data for every selected variable data in Iran. Then, it drew the time series for the data sets. Selected indicators to measure the variables, both endogenous and exogenous, are listed in Table 5 according to the database of the official statistics of Iran.

5.2. Explaining the phenomenon of sprawl and identifying the degree of influence resulted from its drivers in Shiraz Metropolitan Area The study has conducted a simple analysis to identify drivers. In this analysis, two variables affecting urban sprawl have been examined through multiple linear regression. The results are represented in Table 4 which indicates that 99% of changes in urban sprawl are due to these two factors of per capita land consumption and population growth. That is, these two factors have led to major changes in urban sprawl during the course of these years. Also, according to the beta weight calculated, population growth has greater impacts on the metropolitan sprawl compared to per capita land consumption. The t-test results show the significance of 0.95% for the regression. Also, as shown

For sprawl drivers, interpolation is done by selecting the most correlated complete variable to the selected variable that its time series is needed. Table 6 indicates the most fitted nonlinear regression with independent variables, in order to complete its time series. The most fitted pattern is the regression model the “significance” of which is less than 5% and its “adjusted R-square” is the highest compared to that of other models. The higher “adjusted R-square” shows a greater 11

12

Statistics F

Significance

P-value

Lower 95%

0.049672 0.005597 6.68E-08

0.049699 0.00613 1.07E-07

Upper 95% Lower 95.0% Upper 95.0% 7489.478654 2.4273E-161 0.049671957 0.049699 44.17200959 1.71959E-43 0.005597391 0.00613 8.624490904 1.14044E-11 6.68307E-08 1.07E-07

Beta Coefficient Standard Error t Stat

1.43459E-14 7.17294E-15 3551.449848 3.50516E-57 1.07045E-16 2.01972E-18 1.44529E-14

MS

0.996289879 0.992593524 0.992314034 1.42117E-09 56

Remaining (intercept) 0.049685264 6.63401E-06 Population Growth 0.00586364545 0.000132746 Per Capita Land Consumption 0.00000008708 1.00972E-08

Regression 2 Remaining 53 Total 55

Degree of Freedom SS

Multiple R R-square Adjusted R square Standard error Visit ANOVA

Regression statistics

Table 4 Calculations related to multiple regression between independent variable of urban sprawl and dependent variables of population growth and per capita land consumption in the Shiraz Metropolitan Area from 1956 to 2011. Source: interpolation by the authors based on census data of Statistical Center of Iran (1956–2011) and comprehensive development plans of Shiraz.

B. Bagheri, S.N. Tousi

Cities xxx (xxxx) xxx–xxx

Cities xxx (xxxx) xxx–xxx

B. Bagheri, S.N. Tousi

Fig. 8. Average population growth rate (percent) at different time spans of urban sprawl phenomenon in Shiraz.

Fig. 9. Changes in impact rate (share) of “population” and “per capita land consumption” on urban sprawl phenomenon in Shiraz Metropolitan Area from 1956 to 2011.

1.1.7 Fourth step: Examine the “fit” of the estimated theoretical model to determine how well it models the Urban Sprawl phenomenon in SMA, and develop an empirical model using path analysis; the purpose of path analysis, is to develop causal models (Pearl, 2001). At the end, the final correlation of directs and indirect effects will be derived (Heise, 1969). In the previous step, this path was drawn up as the theoretical model through a review of the theoretical frameworks and analyses by other researchers. The article assessed the suitability of theoretical model using “R2” statistics. These statistics representing the degree of the dependent variable variance (urban sprawl), determines how the theoretical model fits the datasets. Also, it explains the dependent variable. The higher this statistic, the stronger the model will be. The rule of thumb in implementation of path analysis is to remove those variables with no significant beta value or path coefficient at the error level of less than 0.05. Having this done, the study carried out the analysis and changed the theoretical relations using the method of trial and error. Then, the experimental model was derived as shown in Fig. 11

proportion of the changes in dependent variable (the time series of which the article aims to draw based on the dependent variable) resulted from the independent variable. After the generalization of fitted regression model to the dependent variables (drivers/factors influencing urban sprawl in Shiraz Metropolitan Area), their time series were estimated separately. Although statistical relations established between the variables in Table 6 were used to complete the data sets, these statistical relations are also provided in the theoretical basis. For example, regarding “per capita land consumption” as a predictor of “creative class”, according to Florida (2005), the theory of creative class believes only in one particular type of human capital as a driver of economic growth of the region. From his perspective, place preference of creative class differs from other citizens; they prefer to live in environments that are of high quality and comfort which has led to the increase of suburbia. 1.1.6 Third step: proposing a theoretically plausible model of urban sprawl phenomenon in the region; drawing the most fitted model of urban sprawl in SMA requires identifying and outlining the theoretical components and drawing theoretical relations. Literature review has identified theoretical components in previous steps. Detecting theoretical relations among identified components is possible using correlation analysis between components as well as supporting documents. In this type of analysis, the path is often represented by a one-way directional arrow that is drawn from an endogenous variable to the exogenous one (Duncan, 1975; Land, 1969; Wright, 1960). The results derived from analyses show a theoretical model represented as Fig. 10. Next step is to test the theoretical model using data gathered and simulated from the SMA

Based on the results summarized in Table 7, after removing insignificant theoretical relations, our derived empirical model could explain 99.7% of urban sprawl phenomenon in SMA, which represents a strong experimental model. Bellow, only direct model table of sprawl has been shown and other 12 indirect model tables are available in appendix. Also, total direct and indirect effects of identified independent variables on the dependent variables of urban sprawl in SMA were calculated (Table 8). Direct effect is the impact of variable X on variable Y. While indirect effect of X on Y is measured through other predictor 13

14

Population Growth

Knowledge economy

Local budgets

Industrialization

Demand for more living space

Speculations

Drivers

Population growth in the city of Shiraz in different years

income in different years (IRR) • Municipal costs in different years (IRR) • Municipal • Municipal per capita income (IRR) number of active Research & Development • The Units of the province number of employees in Research and • The Development Units

province

of workshops with 100 workers and more, • Number regarding the number of employees and the

province

share of the workshops with 10 to 100 workers • The regarding the number of employees and the

room

households and those living in typical • Ordinary residential units regarding the number of rooms households and those living in typical • Ordinary housing units regarding the number of people in a

members

households and those living in typical • Ordinary housing units regarding the number of family

capita real transactions recorded in the • Per registries of deeds and properties number of immovable property transactions • The recorded in the registries of deeds and properties capita real estate transactions registered in the • Per registries of deeds and properties number of movable property transactions • The recorded in the registries of deeds and properties

Type of Measure

Low-rise development

Technological expansions

Transportation Preferences

Network Development

Affordable housing

Drivers

Table 5 List of selected indicators to measure drivers affecting urban sprawl phenomenon in SMA.

phone penetration rate depending on • Mobile the state penetration rate depending on the • Internet state • Proportion of two-stored (and less) buildings

average cost of shipping an urban • The household by province (Rial) of urban households using private • Percent cars in the province

capita highways of the province and city • Per of Shiraz (Unit: km) capita main roads of the province and the • Per city of Shiraz (Unit: km) length of highways of the province and • The the city of Shiraz (Unit: km) length of the main roads of the province • The and the city of Shiraz (Unit: km)

constructions

land ceded by the Housing and Urban • The Development Office for governmental

constructions (Unit: m)

share granted by the Housing and Urban • The Development Office for cooperative

Mehr residential housing units (unit: m)

lands ceded by the Housing and Urban • The Development Office for the construction of

residential units (unit: m)

lands ceded by the Housing and Urban • The Development Office for the construction of

Type of Measure

Land consumption

Tendencies for urbanization

Creative Class Attraction

Government developmental policies

Housing Investment

Purchasing Power and Household Consumption Patterns

Drivers

• Per capita real estate transactions

Rials unit)

of 10 years or more in terms of • Employees literacy level in urban areas of Shiraz employment share of total • Experts employment • Rate of urbanization

Rials unit)

capita annual development budget by the • Per Department of Public Revenue of Fars (1000

function of credits by Urban • The Development Department of Fars (1000

Rials)

average rental price for one square meter • The of housing units in the city of Shiraz (1000

Rials)

Rials)

average price for one square meter of • The housing units in the city of Shiraz (1000

average price for one square meter of • The residential land in the city of Shiraz (1000

an urban household depending on the state

total annual household food • Average expenditure total annual household cost of non• Average food and tobacco of cost housing or share of housing • Average costs from household expenditure

annual income of urban • Average households + average total annual cost of

Type of Measure

B. Bagheri, S.N. Tousi

Cities xxx (xxxx) xxx–xxx

Cities xxx (xxxx) xxx–xxx

B. Bagheri, S.N. Tousi

Table 6 Fitted regression models to interpolate and extrapolate time series of effective drivers of urban sprawl phenomenon in Shiraz Metropolitan Area. Dependent Variable

Independent Variable (Predictive)

Land price Automobile-oriented-ness Expansion of roads network Speculation of land and housing Household income Technological development Local budget Government budget Creative class Knowledge economy Affordable housing Large-scaled industrial units Housing pattern Truncation Municipal income

Model Summarized

Selected pattern as the most fitted

R

R-Square

Adjusted RSquare

Std. Error of the Estimate

F

Sig

Population Conversion rate of lowlands to farms Automobile oriented-ness Expansion of roads network

0.89 0.96

0.79 0.92

0.78 0.92

10.35 6.09

100.20 205.10

0.00 0.00

(cubic) (cubic)

0.95 0.98

0.89 0.97

0.89 0.96

8.99 5.14

150.50 512.24

0.00 0.00

(cubic) (cubic)

Urban sprawl Expansion of roads network Automobile oriented-ness Expansion of roads network Per capita land consumption Automobile-oriented-ness Price of land and housing Population Urbanization Large-scaled industrial units Housing pattern

0.94 1.00 0.83 0.96 0.96 0.92 0.86 0.86 0.91 0.91 0.68

0.89 0.99 0.69 0.92 0.92 0.85 0.75 0.74 0.83 0.83 0.46

0.88 0.99 0.67 0.91 0.92 0.84 0.73 0.72 0.82 0.83 0.44

12.35 1.88 14.22 6.62 9.51 8.30 7.54 3.57 14.68 18.24 18.11

142.31 3089.51 40.28 197.67 205.96 101.49 52.83 50.80 131.85 135.48 23.25

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

(cubic) (cubic) (cubic) (cubic) (cubic) (cubic) (cubic) (cubic) (quadratic) (quadratic) (cubic)

Fig. 10. Theoretically plausible model of urban sprawl in SMA derived from reviewed documents and correlation analysis.

due to the existence of one of the most fundamental weaknesses in urban and regional discussions, the gap between the theory and practice, and this phenomenon is totally unconscious in most metropolitan, therefore there is no study in Iran in the sprawl literature that evaluates its targeted promotion in relation to its positive aspects. A review of the studies carried out in the cities of large, medium and small cities of Iran, in general, on the one hand, the major sprawl factor in large cities of Iran is different in comparing with the sprawl of the Western cities because of increasing the population which results in immigration due to occupational attraction in industrial areas of metropolises, the war between Iran and Iraq in 1980 and the Islamic Revolution of Iran in 1979 (and the adoption of policies to increase the population resulting from it), and the inability of low-income immigrants to find affordable housing within the urban area caused by the land market inflation and the lack of a bank state land for public use on the other. Meanwhile, the incapacity and lack of readiness of management systems in confronting this phenomenon and neglecting infill developments are the cause of

variables. This means, when X is cause acting on Z, the relation between X and Y is an indirect one and Z, in turn, has its own effect on Y. Calculating the general effect of one variable on the other may be carried out through a totalizing of direct and indirect effects. Indirect effects are calculated by multiplying the coefficients of each path.

6. Discussion In spite of many experts' view, Jenks and Burgess (2000), point out a number of sprawl positive aspects such as congestion and fewer crowding, less pollution and more space for services. Although paying attention to the sprawl positive aspects and decentralization of urban parts are merely targeted, and the focus is on the allocation of special sites for this, and this is to some extent taken into account in Western and European societies, but in Iran, sprawl has a completely negative image and avoidance of it is emphasized in the vast majority of studies and projects. Unfortunately, this has been largely unsuccessful in Iran 15

Cities xxx (xxxx) xxx–xxx

B. Bagheri, S.N. Tousi

Fig. 11. Experimental fitted model for the urban sprawl phenomenon in the SMA, with R-square of 99.7%.

Table 7 The results of the analysis in SMA (summarized experimental model after testing the theoretical models). Model

R

R-Square

Adjusted R-Square

1 Model

0.997

1

(constant) Per capita land consumption Population Affordable housing (removed direct relation) Large-scaled industrial units Expansion of roads network Automobile oriented-ness

0.993 0.992 Unstandardized Coefficients B Std. Error 31.543 0.910 0.054 0.020 0.717 0.037 0.010 0.025 − 0.053 0.017 − 0.120 0.034 0.131 0.034

Std. Error of the Estimate 1.72533 Standardized Coefficients Beta – 0.071 0.993 0.011 − 0.083 − 0.163 0.160

t

Sig.

34.659 2.696 19.547 0.401 − 3.135 − 3.543 3.803

0.000 0.009 0.000 0.690 0.003 0.001 0.000

field of sprawl in Iran are focused on the urban level and studied less on the regional level. Also, many studies have mainly sought to prove the existence and recognition of dispersed overlap patterns. Therefore, the exact identification of the factors involved in the occurrence of this phenomenon and the measurement of the impact of each and the relationships between these factors has always been hidden from the views. In addition, the typology of urban sprawl patterns in the region and its investigation in a 55-year period, despite the data constraints in Iran, has made our study in many respects unique to other studies. SMA in Iran, especially after 1990s, has faced linear sprawl up to 15 km distances at its North and North-west. This was coincided with the formation of the new city, Sadra, next to the industrial centers such as electronics and electronic components, trade-service centers such as Persian Gulf Commercial Building, International Trade Fair and University of Payam-Noor, as well as the construction of other new towns such as Golestan town. The region, though, has faced other types of sprawl, including clustering, individual sprawl as well as expansion and extension. According to the analyses, intensity of urban sprawl has a direct relation with the distance from the city center. This necessitates immediate control measures in this area. In this regard, both re-planning of urban development and strategic planning of regional spatial development with the above-mentioned purpose seem to be necessary. Based on the results represented in Table 8, attracting creative class has the most negative effect on urban sprawl after population growth (i.e. an increase in this class results in decreased urban sprawl). According to Fig. 11, developments in knowledge economics as well as communication and IT infrastructures are among the drivers attracting this

the intensification of this phenomenon. In general, the economic and social disorganizations in the city has led to disturbances in the physical development of the city and the end of the natural process and development. Although the factor of population growth and the inability of urban planning and management systems in dealing with this phenomenon has been more or less ineffective in small and medium-sized cities but it is not considered as the main factor. But increasing the per capita consumption of land has been a major sprawl factor in these cities. In many small and medium-sized cities, especially in natural areas such as Sanandaj (Mohammady & Delavar, 2014), Sari and Urumia (Mohammadi et al., 2012), increasing in per capita land use is due to a tendency to marginalization and lack of attention Urban constraints, increasing governmental and communal housing without urban amenities, urban wandering, speculation and activities related to the land trade have led to scattered and unqualified development of city and around the city. In this category of cities, unlike the big cities, the distance to the center of the city has had the least significant in sprawl. In this regard, in the face of this phenomenon, mainly solutions were used such as regional equilibrium (Mohammadi et al., 2012), emphasis on intensive growth (Zali et al., 2016) and vertical development (Mohammady & Delavar, 2014), infill development (Eslami Mahmoudabadi et al., 2013; Zanganeh Shahraki et al., 2011), urban regeneration and brownfields revitalization (Soltani et al., 2017), Urban growth limitation policies such as the green belt and the avoidance of unauthorized construction and marginalization in the suburbs (Dadras et al., 2014). As a remarkable and distinctive feature of this paper, it should be noted that most of the existing studies in the 16

0.071 0.993 − 0.083 − 0.163

0.160

_ _

_ _

_

_

_ _

_ _

_

0.071 3.680 − 0.113 − 0.163

0.293

0.000 0.396

− 2.034 − 0.057

− 0.251

0.032

0.764 0.297

0.039 0.594

0.218

Land Consumption Population growth Industrialization Network Development Transportation Preferences Affordable Housing Purchasing Power and Household Consumption Patterns Creative Class Low-rise Development Knowledge Economy Technological Expansion Speculations Tendencies for Urbanization Local budgets Housing Investment Government Developmental Policies

Direct effects

Total effects

Independent variables

0.15 0.65

0.24

Population growth

17 − 0.78

1.51

0.25

Purchasing Power and Household Consumption

Indirect effects

1.00

− 0.51

Local budgets

0.31

0.34

− 0.24

0.13

Network Development

Table 8 Total direct and indirect effects of independent variables on dependent variables of urban sprawl in SMA.

0.62

Housing Investment

0.76 − 0.61

0.69

Creative Class

0.08

0.62

− 0.18

− 0.30

Speculations

− 1.67

− 1.41

2.69

Government Developmental Policies

0.53 0.31

1.03

Transportation Preferences

− 0.63

0.94 − 0.80

Land Consumption

0.75

Technological Expansion

0.47 0.59

Affordable housing

0.22

0.04 0.59

0.76 0.30

0.03

− 0.25

− 2.03 − 0.06

0.00 0.40

0.13

0.00 2.69 − 0.03 0.00

Total indirect effects

B. Bagheri, S.N. Tousi

Cities xxx (xxxx) xxx–xxx

Cities xxx (xxxx) xxx–xxx

B. Bagheri, S.N. Tousi

Fig. 12. Explanatory theoretical framework of sprawl – proposed by the research.

the provision of infrastructure and activity clusters seem to be influential as they will lead to the selection of residents with environmental responsibility. On the other hand, government policies in recent years aimed at construction of cheap residential centers in cheap peripheral lands. This has attracted low-income residents from other neighboring towns and villages to the suburbs which resulted in increased land-use change of farmlands and horticultural lands. Thus, it must be prioritized to stop or to effectively lead the programs planned in order to provide housings for low-income classes through Mehr housing program-which mostly occurs in the suburbs as leapfrog sprawl-towards brownfield regeneration and urban distressed areas that have the best accessibility and service-communication infrastructures. For this purpose, it may be beneficial to consider the theories of sustainable developments, compact city, smart growth, growth management and new urbanism. The reason is that they put more spatial and physical dimensions on the agenda when dealing with urban sprawl.

class. The other main drivers are speculation of land and housing, speculators and increased land price. Speculation of land and housing is one of the drivers influencing this phenomenon through increased per capita land consumption. Other drivers also affected this phenomenon in SMA, including automobile usage which has resulted in automobileoriented urban development, increased urban communication networks and ultimately stimulation of urban dispersion in valuable natural and rural areas. Therefore, considering neighborhood-oriented development models, developments based on public transport and pedestrian-oriented developments should be considered in urban development plans. The study of drivers affecting urban sprawl in SMA, suggests that unlike many reviewed experiences, especially in the context of American cities, increased per capita land consumption is not the major factor influencing urban sprawl. Instead, this article suggests that population growth and the lack of coherent and purposeful planning are the main causes. Population growth was caused by job opportunities, higher household income as well as development policies of affordable housing around SMA. In this regard, applying policies in order to restrict immigrations to SMA and leading them indirectly to supplement pre-planned centers seem to be effective. Obviously, exiting current situation requires planned residentialactivity clusters such as satellite cities in farther distances from the metropolitan of Shiraz. Also, there is a need to facilitate polycentric development based on supplement centers like Marvdasht. As Modarres (2011) in his study acknowledges that polycentric urban employment patterns may provide a better explanation of commuting patterns. In this regard, establishment of regional networks, strengthening competitiveness, and increasing attractions of center number two in the region, are considered as vital approaches. Also, according to the results, in terms of incidence, after “population growth” (with direct and positive impact on the occurrence of the phenomenon), “creative class absorption” has also a high impact on sprawl which is in the opposite direction; that is an increase in this class leads to the decrease in sprawl. Also, the developments of knowledge economy and ICT infrastructure are among the main factors attracting this class. Another contributing factor is “speculation in land and housing, mediation and rising prices”. Land and housing speculation is the other contributing factor through increasing the per capita consumption of land. Other factors also affect occurrence of this phenomenon in SMA including automobile-orientedness which leads to the city development based on cars and increased length of the communication networks and, finally, the stimulation of Urban Sprawl in the areas of precious natural and rural surrounding. Therefore, considering the neighborhood-oriented development models, pedestrian-oriented development and transit-oriented development should be on the urban development agenda. Drivers like those derived from experimental model and also their priorities in transposition program as well as resource and budget allocations. Therefore, according to the fitted empirical model, prioritizing policies for attracting creative class as well as

7. Conclusion As the analytical summary of the research, different from the results of other researches and as a result of the studies conducted by the authors, it is obvious that conventional policies to deal with sprawl which are purely physical- spatial, such as smart growth and infill development aimed to reduce per capita land consumption before addressing the aspatial drivers, are in fact inefficient and unsustainable and may idealistically lead to fix the problem periodically. Factors such as government's centralized economic-based policies that lead to the concentration of facilities and activities in one city, formation of growth poles, as well as deprivation and underdevelopment of the other centers, on one hand and the willingness of society to city life on the other hand, are among the fundamental factors shaping urban sprawl through increasing population growth in urban areas. Finally, dealing with factors such as improper and unsustainable consumption and production must be considered before physical methods, which are conducted both by the private sectors and people in the form of unsustainable behavior (such as reliance on cars) and by the government which leads to incorrect placement for new urban developments in cheap natural surrounding areas through inefficient management. Based on simplified conceptual diagram below as a result of summing up the analytical results of the paper, in addition to the aspatial drivers and government macroeconomic policies, global technological advances also lead to the intensify this phenomenon through facilitating decentralization of urban centers of residence and activity to the periphery. Explanatory theoretical framework proposed by this research (Fig. 12) provides a guideline to prioritize the policies dealing with the phenomenon of sprawl through classification of the factors into three categories of input- output, outcome. That is, in order to deal with the sprawl, it is necessary to provide solutions for the aspects of aspatial 18

Cities xxx (xxxx) xxx–xxx

B. Bagheri, S.N. Tousi

urban fringe from a spatio-temporal perspective. Applied Spatial Analysis and Policy, 9(2), 233–250. Fischel, W. (1982). The urbanization of agricultural land: A review of the National Agricultural Lands Study. Land Economics, 58(2), 236–259. Flew, T. (2012). Creative suburbia: Rethinking urban cultural policy - the Australian case. International Journal of Cultural Studies, 231–246. Florida, R. (2002a). The rise of the creative class. New York: Basic Books. Florida, R. (2002b). The rise of the creative class and how it's transforming work, leisure, community and everyday life. New York: Basic Books. Florida, R. (2005). Cities and the creative class. New York: Routledge. Frumkin, H. (2002). Urban sprawl and public health. Public Health Reports, 117, 201–217. Fulton, W., Pendall, R., Nguyen, M., & Harrison, A. (2001). Who sprawls the most? How growth patterns differ across the US. Washington, DC: The Brookings Institution Center on Urban and Metropolitan Policy. Galster, G., Hanson, R., Ratcliffe, M., Wolman, H., Coleman, S., & Freihage, J. (2001). Wrestling sprawl to the ground: Defining and measuring an elusive concept. Housing Policy Debate, 12(4), 681–717. Geo-Kommunikation, L. G. (2000–2002). Lexikon der Geowissenschaften. Heidelberg: Berlin. Giuliano, G. (1989). Literature synthesis: Transportation and urban form. Report prepared for the Federal Highway Administration under Contract DTFH61-89-P-00531. Glaeser, E., Kahn, M., & Chu, C. (2001). Job sprawl: Employment location in U.S. metropolitan area. Washington, DC: The Brookings Institution. Gordon, P., & Richardson, H. (1997). Are compact cities a desirable planning goal? Journal of the American Planning Association, 63(1), 95–106. Gouda, A. A., Hosseini, M., & Masoumi, H. E. (2016). The status of urban and suburban sprawl in Egypt and Iran. GeoScape, 10(1), 1–15. Grimm, N., Grove, J., Pickett, S., & Redman, C. (2000). Integrated approaches to longterm studies of urban ecological systems. Bioscience, 50, 571–584. Guhathakurta, S., & Gober, P. (2010). Residential land use, the urban Heat Island, and water use in phoenix: A path analysis. Journal of Planning Education and Research, 30(1), 40–51. Hajizadeh, M., Campbell, M., & Sarma, S. (2015). A spatial econometric analysis of adult obesity: Evidence from Canada. Applied Spatial Analysis and Policy. http://dx.doi.org/ 10.1007/s12061-015-9151-5. Harvey, R. O., & Clark, W. A. V. (1965). The nature and economics of urban sprawl. Land Economics, 41(1), 1–9. Hasse, J., & Lathrop, R. (2003a). A housing-unit-level approach to characterizing residential sprawl. Photogrammetric Engineering and Remote Sensing, 69(9), 1021–1030. Hasse, J., & Lathrop, R. (2003b). Land resources impact of urban sprawl. Applied Geography, 23, 159–175. Heise, D. (1969). Problems in path analysis and causal inference. Sociological Methodology, 1, 38–73. Holdren, J. (1991). Population and the energy problem. Population and Environment, 12, 231–255. Hosseini, S. A., Zanganeh Shahraki, S., Farhudi, R., Hosseini, S. M., Salari, M., & Pourahmad, A. (2014). Effect of urban sprawl on a traditional water system (qanat) in the City of Mashhad, NE Iran. Urban Water Journal, 7(5), 309–320. http://dx.doi.org/ 10.1080/1573062X.2010.484497. Inostroza, L., & Helena Tábbita, J. (2016). Informal urban development in the greater Buenos Aires area: A quantitative-spatial assessment based on households' physical features using GIS and principal component analysis. Procedia Engineering, 161, 2138–2146. Jacquin, A., Misakova, L., & Gay, M. (2008). A hybrid object-based classification approach for mapping urban sprawl in periurban environment. Landscape and Urban Planning, 84, 152–165. Jaeger, J., Bertiller, R., Schwick, C., Cavens, D., & Kienast, F. (2010). Urban permeation of landscapes and sprawl per capita: New measures of urban sprawl. Ecological Indicators, 10(2), 427–441. Jat, M., Garg, P., & Khare, D. (2008). Monitoring and modelling of urban sprawl using remote sensing and GIS techniques. International Journal of Applied Earth Observation and Geoinformation, 10, 26–43. Jenks, M., & Burgess, R. (2000). Compact cities: Sustainable urban forms for developing countries. 2004. Taylor & Francis e-Library. Johnson, L. (1991). Advances in telecommunications technologies that may affect the location of business activities. Rand Note N–3350–SF, RAND Corporation. Kahn, M. (2001). Does sprawl reduce black/white housing consumption gap? Housing Policy Debate, 12, 77–86. Kaur, P., & Singh, R. (2005). Conflict resolution in urban and rural families: A factor analytical approach. The Journal of Business Perspective, 9(1), 59–67. Kunstler, J. (1993). The geography of nowhere. New York: Touchstone Books. Land, K. (1969). Principles of path analysis. Sociological Methodology, 1, 3–37. Retrieved from http://www.jstor.org/stable/270879. Landscape Gesellschaft fur Geo-Kommunikation (2000–2002). Lexikon der Geowissenschaften. Vol. 5. Heidelberg, Berlin: Spektrum Akademischer Verlag. Leser, H., & Huber-Frohli, J. (1997). In H. Leser (Ed.). Diercke-Worterbuch allgemeine Geographie (pp. 1037). Munchen: estermann, Braunschweig, and Deutscher Taschenbuch Verlag. Masoumi, H. E. (2012a). A new approach to the Iranian urban planning, using neo-traditional development. PhD dissertationTechnical University of Dortmund, Germany, Faculty of Spatial Planning. Masoumi, H. E. (2012b). Urban sprawl in iranian cities and its differences with the western sprawl. SPATIUM International Review, 27, 12–18. http://dx.doi.org/10. 2298/SPAT1227012E. McGahey, R., Malloy, M., Kazanas, K., & Jacobs, M. (1990). Financial services, financial centers: Public policy and the competition for markets, firms and jobs. Boulder, CO:

drivers, national and international issues that requires collective determination. It seems that, modifying infrastructures during the consequent steps, causes of aspatial and incomplete faulty cycles of population growth in metropolitan areas will be corrected in favor of sustainable urban development. However, experimental studies will be needed to confirm this relation. References Acioly, C., & Davidson, F. (1996). Density in urban development. Building Issues, 8(3), 3–25. Akademie fur Raumforschung und Landesplanung (1970). Handworterbuch der Raumforschung und Raumordnung. Band III. Hannover: Gebruder Jarnecke Verlag3974. Akademie fur Raumforschung und Landesplanung (ARL), & Schweizerische Vereinigung fur Landesplanung (VLP) (1999). Deutsch-Schweizerisches Handbuch der Planungsbegriffe. Hannover: Verlag der Akademie fur Raumforschung und Landesplanung. Al Gore (December 1998). Speech to Democratic Leadership Council Annual Conference. Retrieved from www.cnu.org/inthenews.html. Anselin, L. (1995). Local indicators of spatial association – LISA. Geographical Analysis, 27(2), 93–115. Asgarian, A., Jabbarian Amir, B., Alizadeh Shabani, A., & Feghhi, J. (2014). Predicting the spatial growth and urban sprawl in sari, Iran using Markov cellular automata model and Shannon entropy. Iranian Journal of Applied Ecology, 2(6), 13–25. Aurand (2013). Does sprawl induce affordable housing? Growth and Change: A Journal of Urban and Regional Policy, 44(4), 631–649. Barnes, K., Morgan, J., Roberge, M., & Lowe, S. (2001). Sprawl development: Its patterns, consequences, and measurement. A white paperTowson University. Batty, M., Xie, Y., & Sun, Z. (1999). Modeling urban dynamics through GIS-based cellular automata. Computers, Environment and Urban Systems, 23(3), 205–233. Beck, R., Kolankiewicz, L., & Camarota, S. (2003). Outsmarting smart growth: Population growth, immigration, and the problem of sprawl. Washington, DC: Center for Immigration Studies. Bekele, H. (2005). Urbanization and urban sprawl. Master of Science ThesisKungliga Tekniska Hogskolan: Department of infrastructure section of building and real estate economics. Benfield, F., Raimi, M., & Chen, D. (1999). Once there were Greenfields: How urban sprawl is undermining America's environment. The Natural Resources Defense Council, Washington, DC: Economy and Social Fabric. Berry, D., & Plaut, T. (1978). Retaining agricultural activities under urban pressures: A review of land use conflicts and policies. Policy Sciences, 9, 153–178. Bertaud, A., & Malpezzi, S. (1999). The spatial distribution of population in 35 world cities: The role of markets, planning and topography. Madison: The University of Wisconsin. Bhatta, B. (2009). Modelling of urban growth boundary using geoinformatics. International Journal of Digital Earth, 2(4), 359–381. Bhatta, B. (2010). In S. Balram, & S. Dragicevic (Eds.). Analysis of urban growth and sprawl from remote sensing data. Berlin Heidelberg: Springer. Boyce, R. (1963). Myth versus reality in urban planning. Land Economics, 39(3), 241–251. Brueckner, J., & Kim, H. (2003). Urban sprawl and the property tax. International Tax and Public Finance, 10, 5–23. Burchell, R., & Galley, C. (2003). Projecting incidence and costs of sprawl in the United States. Transportation Research Record, 1831, 150–157. Carnes, A., & Ogneva-Himmelberger, Y. (2012). Temporal variations in the distribution of West Nile virus within the United States; 2000–2008. Applied Spatial Analysis and Policy, 5, 211–229. http://dx.doi.org/10.1007/s12061-011-9067-7. Cervero, R. (2003). Road expansion, urban growth, and induced travel. Journal of the American Planning Association, 69(2), 145–163. Cheng, J., & Masser, I. (2003). Urban growth pattern modeling: A case study of Wuhan city, PR China. Landscape and Urban Planning, 62, 199–217. Clawson, M. (1962). Urban sprawl and speculation in suburban land. Land Economics, 38(2), 99–111. Conley, J., Stein, R., & Davis, C. (2014). A spatial analysis of the neighborhood scale of residential perceptions of physical disorder. Applied Spatial Analysis and Policy. http://dx.doi.org/10.1007/s12061-013-9099-2. Dadras, M., ZulhaidiMohd Shafri, H., Ahmad, N., Pradhan, B., & Safarpour, S. (2014). Land use/cover change detection and urban sprawl analysis in Bandar Abbas City, Iran. The Scientific World Journal, 1–12. http://dx.doi.org/10.1155/2014/690872. Duncan, O. (1975). Introduction to structural equation models. New York: Academic Press. Earth Systems Research Institute (2009). ArcGIS Desktop 9.3 Help. Retrieved from http:// webhelp.esri.com/arcgisdesktop/9.3/index.cfm?TopicName=How%20Average %20Nearest%20Neighbor%20Distance%20(Spatial%20Statistics)%20works. Ermer, K., Mohrmann, R., & Sukopp, H. (1994). In K. Buchwald, & W. Engelhardt (Eds.). Stadt und Umwelt (Vols. 12 des Handbuches “Umweltschutz – Grundlagen und Praxis”). Bonn: Economica Verlag. Eslami Mahmoudabadi, S., Soroushnia, R., & Zekri, A. (2013). Evaluating Iranian urban development plants in order to develop a development plan for Sirjan City of Iran. Technical Journal of Engineering and Applied Sciences, 3(19), 2365–2370. Ewing, R. (1997). Is Los Angeles-style sprawl desirable? Journal of the American Planning Association, 63(1), 107–126. Ewing, R., Pendall, R., & Chen, D. (2002). Measuring sprawl and its impact. 1. Washington, DC: Smart Growth America. Feng, L., Du, P., Zhu, L., Luo, J., & Adaku, E. (2016). Investigating sprawl along China's

19

Cities xxx (xxxx) xxx–xxx

B. Bagheri, S.N. Tousi

Soltani, A., Hosseinpour, M., & Hajizadeh, A. (2017). Urban sprawl in Iranian mediumsized cities; investigating the role of masterplans. International Journal of Sustainable Development, 10(1), 122–131. Stone, B., Jr. (2008). Urban sprawl and air quality in large US cities. Journal of Environmental Management, 86, 688–698. Straus, J., Chang, H., & Hong, C.-y. (2016). An exploratory path analysis of attitudes, behaviors and summer water consumption in the Portland Metropolitan Area. Sustainable Cities and Society, 23, 68–77. Sturm, R., & Cohen, D. (2004). Suburban sprawl and physical and mental health. Public Health, 118, 488–496. Sudhira, H., Ramachandra, T., & Jagadish, K. (2004). Urban sprawl: Metrics, dynamics and modelling using GIS. International Journal of Applied Earth Observation and Geoinformation, 5(1), 29–39. Terzi, F., & Bolen, F. (2011, January). The potential effects of spatial strategies on urban sprawl in Istanbul. Urban Studies, 49(6), 1229–1250. Tewolde, M., & Cabral, P. (2011). Urban sprawl analysis and modeling in Asmara, Eritrea. Remote Sensing, 3, 2148–2165. Theil, H. (1967). Economics and information theory. Amsterdam: North-Holland. Thomas, R. (1981). Information statistics in geography: Geo abstracts. United Kingdom, Norwich: University of East Anglia. Torrens, P. (2008). A toolkit for measuring sprawl. Applied Spatial Analysis and Policy, 1, 5–36. http://dx.doi.org/10.1007/s12061-008-9000-x. Torrens, P., & Alberti, M. (2000). Measuring sprawl. Centre for Advanced Spatial Analysis, University College London. Tsai, Y. (2005). Quantifying urban form: Compactness versus sprawl. Urban Studies, 42, 141–161. USHUD (1999). The state of the cities: Third annual report. Washington, DC: US Department of Housing and Urban Development. Wang, W., Zhu, L., Wang, R., & Shi, Y. (2003). Analysis on the spatial distribution variation characteristic of urban heat environmental quality and its mechanism—A case study of Hangzhou city. Chinese Geographical Science, 13(1), 39–47. Weng, Q., Liu, H., & Lu, D. (2007). Assessing the effects of land use and land cover patterns on thermal conditions using landscape metrics in city of Indianapolis, United States. Urban Ecosystem, 10, 203–219. Wright, S. (1960). Path coefficients and path regressions: Alternative or complementary concepts? Biometrics, 16, 189–202. Yang, X., & Lo, C. (2003). Modelling urban growth and landscape changes in the Atlanta metropolitan area. International Journal of Geographical Information Science, 17, 463–488. Yanos, P. (2007). Beyond “landscapes of despair”: The need for new research on the urban environment, sprawl, and the community integration of persons with severe mental illness. Health & Place, 13, 672–676. Yeh, A., & Li, X. (2001). Measurement and monitoring of urban sprawl in a rapidly growing region using entropy. Photogrammetric Engineering & Remote Sensing, 67(1), 83–90. Yigitcanlar, T. (2007). The making of urban spaces for the knowledge economy: Global practices in proceedings the 2nd international symposium on knowledge cities: Future of cities in the knowledge economy. Malaysia: Queensland University of Technology. Zali, N., Hashemzadeh Ghal'ejough, F., & Esmailzadeh, Y. (2016). Analyzing urban sprawl of Tehran metropolis in Iran (during 1956-2011). Anuário do Instituto de Geociências, 39(3), 55–62. http://dx.doi.org/10.11137/2016_3_55_62. Zanganeh Shahraki, S., Sauri, D., Serra, P., Modugno, S., Seifolddini, F., & Pourahmad, A. (2011). Urban sprawl pattern and land-use change detection in Yazd, Iran. Habitat International, 35, 521–528. http://dx.doi.org/10.1016/j.habitatint.2011.02.004. Zhang, X., Chen, J., Tan, M., & Sun, Y. (2007). Assessing the impact of urban sprawl on soil resources of Nanjing city using satellite images and digital soil databases. Catena, 69, 16–30.

Westview Press. Mitchell, J. (2001). Urban sprawl: The American dream? National Geographic, 200(1), 48–73. Modarres, A. (2011). Polycentricity, commuting pattern, urban form: The case of Southern California. International Journal of Urban and Regional Research, 35(6), 1193–1211. http://dx.doi.org/10.1111/j.1468-2427.2010.00994.x. Moe, R. (1999). The sprawling of America – Federal policy is part of the problem; can it be part of the solution? Washington, D.C.: 22 January address to: the National Press Club. Mohammadi, J., Zarabi, A., & Mobaraki, O. (2012). Urban sprawl pattern and effective factors on them: The case of Urmia city, Iran. Journal of Urban and Regional Analysis, 4(1), 77–89. Mohammadian Mosammam, H., Tavakoli Nia, J., Khani, H., Teymouri, A., & Kazemi, M. (2016). Monitoring land use change and measuring urban sprawl based on its spatial forms, the case of Qom city. Egyptian Journal of Remote Sensing and Space Science. http://dx.doi.org/10.1016/j.ejrs.2016.08.002 (article in press). Mohammady, S., & Delavar, M. R. (2014). Urban sprawl modelling. The case of Sanandaj City, Iran. Journal of Settlements and Spatial Planning, 5(2), 83–90. Nelson, A. (1990). Economic critique of prime farmland preservation policies in the United States. Journal of Rural Studies, 6(2), 119–142. Nelson, A. (1999). Comparing states with and without growth management: Analysis based indicators with policy implications. Land Use Policy, 16, 121–127. Nelson, A., & Duncan, J. (1995). Growth management principles and practices. Chicago: The American Planning Association. Newman, P., & Kenworthy, J. (1988). The transport energy trade-off: Fuel-efficient traffic versus fuel-efficient cities. Transportation Research, 22(3), 163–174. Oueslati, W., Alvanides, S., & Garrod, G. (2015). Determinants of urban sprawl in European cities. Urban Studies, 52(9), 1594–1614. Pearl, J. (2001). Direct and indirect effects. Proceedings of the seventeenth conference on uncertainy in artificial intelligence (pp. 411–420). San Francisco, CA: Morgan Kaufmann. Peiser, R. (1989). Density and urban sprawl. Land Economics, 65, 193–204. Pendall, R. (1999). Do land-use controls cause sprawl? Environment and Planning, 26(4), 555–571. Porat, I., Shoshany, M., & Frenkel, A. (2012). Two phase temporal-spatial autocorrelation of urban patterns: Revealing focal areas of re-urbanization in Tel Aviv-Yafo. Applied Spatial Analysis and Policy, 5, 137–155. http://dx.doi.org/10.1007/s12061-0119065-9. Riitters, K., O'Neill, R., Hunsaker, C., Wickham, J., Yankee, D., Timmins, S., ... Jackson, B. (1995). A factor analysis of landscape pattern and structure metrics. Landscape Ecology, 10(1), 23–39. Roshan, G., Zanganeh Shahraki, S., Sauri, D., & Borna, R. (2010). Urban sprawl and climatic changes in TEHRAN. Journal of Environmental Health Science & Engineering, 7(1), 43–52. Sanders, P., Zuidgeest, M., & Geurs, K. (2015). Liveable streets in Hanoi: A principal component analysis. Habitat International, 49, 547–558. Savitch, H. (2003). How suburban sprawl shapes human well-being. Journal of Urban Health: Bulletin of the New York Academy of Medicine, 80(4), 590–607. Scott, A. J. (2006). Creative cities: Conceptual issues and policy questions. Journal of urban affairs, 28(1), 1–17. Shariat-Mohaymany, A., & Shahri, M. (2016, June 27). Crash prediction modeling using a spatial semi-local model: A case study of Mashhad. Iran Application Spatial Analysis. http://dx.doi.org/10.1007/s12061-016-9199-x. Siedentop, S., & Fina, S. (2012). Who sprawls most? Exploring the patterns of urban growth across 26 European countries. Environment and Planning, 44, 2765–2784. Sierra Club (1999). The dark side of the American Dream: The costs and consequences of sunburban sprawl. San Francisco: CA. Retrieved from http://www.sierraclub.org. Slack, E. (2002). Municipal finance and the pattern of urban growth. Urban papers.

20