Urban Growth Structure and Travel Behavior in Tehran City

This paper investigates major characteristics of Tehran’s urban growth structure, how various land use factors such as “density”, “diversity”, “design” and “accessibility” affect travel behavior, population growth and land use development, and future travel demands. Tehran city is currently developing in ways that are likely to increase sprawl and automobile-dependency, which increase problems including traffic and parking congestion, consumer costs, traffic accidents, pollution emissions and inadequate mobility for non-drivers. This analysis indicates that the growth management policies in Tehran’s Comprehensive Plan can significantly reduce vehicle travel and associated problems, resulting in a more sustainable urban development path. This information is useful for evaluating the ability of policies such as Smart Growth, New Urbanism and Accessibility Management to help achieve transport-land use planning objectives.


Introduction
Today, most urban residents frequently travel from their residential neighborhood to other areas for activities such as shopping, work and recreation.As a result, transport systems and land-use systems interact in many ways (Black, 2003;Hanson and Giuliano 2004;Rodrigue et al., 2006).
Until the early twentieth century, most cities, including Tehran city, developed in the form of compact and mixed use neighborhoods where most commonly-used services were located within walking distance of homes.However, increased motor vehicle travel significantly changed this urban development pattern.After the World War II and modernization in Iran, due to the rapid growth of using automobiles and following modern land use regulations, a new urban development pattern together with the significant growth of cities became widespread in this country (Farnaz Arefian & Iradj Moeini, 2016).Today, this kind of pattern is called "urban sprawl" which encourages and intensifies spatial separations of different locations for activities and increases private automobile use (Frumkin et al. 2004;Bruegmann, 2005;Morris, 2005;Burchell et al., 2005;Bhatta, 2010;Williamson, 2010).
From 1970s on, many planning experts, particularly in Europe and North America, began supporting and theorizing new approaches of urban design and planning in order to help reduce negative influences (including traffic congestion, air pollution and urban sprawl) of using automobiles in a widespread manner (Keats, 1958;Mumford, 1964;Leavitt, 1970;Mitchell, 1970;Schneider, 1972;Buel, 1972).Their basic idea is to encourage more compact and integrated urban development, so non-motorized transportation (also known as "Active Transportation" and "Human Powered Transportation") and public transportation could serve a greater portion of trips.Such approaches are now known as "Compact Cities," "New Urbanism," "Transit Oriented Development" and "Smart Growth" (Jacobs, 1961;Kunstler, 1993;Calthorpe, 1993;Jenks and Burgess, 2000;Duany and et al., 2000;Calthorpe and Fulton, 2001;Hayden, 2003;Talen, 2005;Grant, 2006).They all reflect same basic idea: land development patterns affect people's travel behaviors, and as a result, changing travel patterns can help reduce many of the problems resulting from motor vehicle traffic.

Problem Definition
Tehran City's current policies result dispersed and automobile dependent development.As extensively explored in scholarly research, sprawl and automobile-based planning causes a variety of economic, social and environmental problems as summarized in Table 1.

Aims and Research Questions
This article is motivated by an attempt to analyze urban growth structure as a tool for controlling land use expansion and accessibility-based planning in the Tehran city.Our investigation intends to deal with many of the land use and transportation factors, which still appear to be underrepresented in growth management research in the Tehran.
The article is based on the following main objectives:

•
Characterizing the urban growth structure in the Tehran and analyzing the urban growth aspects in the land use and transportation.
• Estimate of urban growth quality.

•
Survey of public values and attitudes about car use/ownership, travel mode choice and Preferences for living in smart or sprawl neighborhoods.

•
Measuring of land use factors impacts on Vehicle Miles Traveled (VMT) for Tehran city.

A Brief Review of Literature
Land use and transport development is best understood as a product of socio-economic development.The complexity of this connection has undergone great shifts, since it is determined by local, regional, state, nationwide, and even global issues.
After the domination of "sustainability" talks in urban studies and planning, new regulations in transport and air pollution have caused significant effects on policy decisions made in land use and urban sustainable development achievements (VTPI, 2005;Department for Transport, 2007;Schiller et al. 2010;Black, 2010;Bauer, 2010;SUMP, 2013;Taniguchi, 2015;Gudmundsson et al. 2016;www.epa.gov).One of the reasons that LUTI has attracted planners, designers, transport engineers and geographers is that both transportation and land use policies absorb significant investments, which can have key influences on urban development patterns' change and evolution.Source: Authors' work.
However, existing LUTI models are unable to forecast the impact of future urban-policy responses, accessibility impacts land values and shapes the location behavior of households and firms, which in turns impacts observed patterns of spatial interactions.Thus, to adequately assess and evaluate the long-term impacts of investment and policies affecting land use on transport and vice versa, a more robust methodology is needed for deriving accessibility indices as the feedback mechanism of the land-use-transport link.(Wegener 2011in Acheampong andSilva, 2015: 27-28).The potential impacts of individual lifestyles and preferences are another important aspect that taken into consideration in this study.
Hence, this article discusses the key components that have constituted the focus of LUTI research and operational model development based on a conceptual framework shown in Figure 1.As shown in the conceptual framework, travel behavior is influenced by urban growth structure via accessibility, travel demand and dwelling demand factors as well as cognitive factors such as attitudes, preferences, perception, etc.

Materials and Methods
In this research, a case study method has been used to achieve the objectives.The case study as a research method can include various techniques.Among methods, we used a questioner to survey attitudes and preferences from people in the area.In addition, secondary data were used throughout the whole study by using SPSS Statistical Software and spatial/geographical distribution models such as Gini coefficient, Entropy coefficient, Shannon entropy, Holdren model and TOD index (TOD index method in this study suggested by authors base on Gini coefficient rationality).These data were collected at "Iran Statistics Center", "Tehran Municipality's Comprehensive Plan" and "Tehran Comprehensive Transportation and Traffic Company".

Study Area and Location
Tehran is the largest metropolitan in Iran and serves as a capital city of the country.It has 22 urban districts and each district has its own municipality, all of which work under the head municipality (Figure 2).

Population
Urban growth is most directly determined by a growing population.Since 1926 to 2015, its population has always had an ascending trend due to modernization, centralization and domestic immigrations in such a manner that the figure has reached from 250 thousands to 8.7 millions in the 90-year period (http://salnameh.sci.org.ir)(Figure 3).

Mobility Condition
Urban growth and land use patterns affect accessibility (people's ability to reach desired services and activities), which affects mobility (the amount and type of travel activity).Automobile ownership per capita is 0.38 cars per urban resident.More than 4.5 million automobiles and more than 3 million motorcycles operate in Tehran.On daily basis, about 10.3 million liters of gasoline, 3.9 million liters of gas oil and 1.3 million cubic meters of CNG are consumed in the city.The highest share of motor vehicles in urban daily travels belongs to automobiles (private cars) and the rate of inner-city trips is increasingly augmenting (Tehran Comprehensive Transportation and Traffic Co., 2015).
In the current years, the number of daily trips has increased from 14 million to nearly 18 million in the metropolitan region.This results in part from the increase in the distance between origins and destinations, as well as increase in automobile ownership rate.In other words, land use-transport policies have mostly focused on "mobility" rather than "accessibility" (Figure 4).

Modes of transportation
Most development from 1950 and 2015 was automobile dependent, designed primarily for automobile access with little consideration for other modes (Figure 5 and Figure 6).Then, the statistic data related to the share of vehicles in daily trips in Tehran shows that the highest figure belongs to private cars (44%) and after that to taxi cabs (22%), revealing the fact that the main mode of transportation is motor-based and the dominant pattern of urban growth is vehicle-based (Figure 7 and Figure 8).The share different vehicles has in traffic congestion reveals that private cars have the most influence on traffic congestion in such a way that they cause 65.5% of urban traffic congestion.After private cars, different types of taxi cabs create 20% of urban traffic congestion.

Automobile-Dependency
The development of the highway system and car ownership/use, which were driven by municipal regulations, has been a great influence on growth structure.Pearson Correlation among variables including motor vehicle ownership, parking area, motor transportation network area, time of trips and speed of trips totally indicates significant correlation, meaning that the pattern of land use-transport development in all districts has moved along the condition of automobile dependency (Table 3).

Mobility Condition by Drawing on Geographical Distribution
The geographical distribution of mobility condition in all districts of Tehran is shown in the following figure in such a manner that moving from central business districts (CBDs) toward peripheral districts; one can see more motor transport network area, parking area, accident rates and automobile ownership rate.In contrast, moving from peripheral districts toward CBDs, one can see more use of public transportation and more non-motor transportation/walking (Figure 9).

Estimate of Urban growth quality
According to conceptual framework in this article, the geographical distribution of population and land use determines the location of activities.The need for interaction arises as a consequence of the spatial separation between population and land-use activities.The transport system creates opportunities for interaction or mobility, which can be measured as accessibility.The distribution of accessibility in urban space, over time, co-determines location decisions and so results in changes in the urban growth quality.
According to results of table 6, Tehran city have moved toward a sprawl quality more than compact quality.But it should be noted that the overall growth determined by increasing in city population.

Survey of Public Values and Attitudes
According to the questionnaire survey on individual tendency to car ownership/use, preferences for travel mode choice, preference for living in sprawl/smart neighborhoods in samples (n=300; samples: 50% female/ 50% male; from 20-65 years old); one can say that travel behavior /demand and choose locations/neighborhoods in respondents is briefly toward car ownership/use and living in sprawl neighborhoods (Figure 10).
Tendency to car ownership and use (%) Preferences for travel mode choice (%)

Measuring of Land Use factors Impacts on VMT
During past half century, transportation planners have mostly used traffic models for assessing demands and the effects of changes in transport systems.Those models use land use factors as an output.However, they are not sensitive to many land use factors emphasized in the current study (Frank and Pivo 1994;USEPA 2002;Hunt, 2001).During past decades many attempts have been made to improve these models in order to better evaluate how land use factors will affect future travel activity (Cervero and Kockelman, 1997;Rosenbaum and Koenig, 1997;Buch and Hickman, 1999;Cambridge Systematic, Inc. 1994;Dunphy and Fisher, 1996;Ewing and DeAnna and Li, 1996;Frank and Pivo, 199;USEPA, 2002).Many studies indicate that urban growth structural variables (i.e., density, diversity, design, destination accessibility, and distance to transit) have statistically significant influence on travel behavior (e.g., Aditjandra, Mulley & Nelson 2013, Grunfelder & Nielsen 2012, Gim 2013, Handy, Cao & Mokhtarian 2005, Meurs & Haaijer 2001, Naess 2013).
Here, we have used an equation for definition and then estimating the effects of changes in land use and urban design on travel demand (equation 1).It uses the extent of inclination of some factors including Density, Diversity, Design and Destinations depending on the features of built environment, neighborhood units, regional access and the amount of produced trips in a neighborhood unit.Those factors are used (for example in econometrics) for calculating the percentage of changes in total Vehicle Trips (VT) and VMT as a result of different land uses and urban designs.After preparing required data in this regard (table 6), the estimation of different factors of density, diversity, design and destination could be done (table 7).Regarding to the changes in four indexes of "density", "diversity", "design" and "accessibility" in perspective of 2025, their effects on VMT calculated above, one can say that the total VMT per capita will decrease form 11.45 km in 2015 to 7.84 km in 2025 in daily basis.The reduction in daily VMT per capita has deep consequences on urban growth patterns.Such reductions can help decease various negative consequences caused by motor vehicle travel and the road and parking facilities it requires, including land consumption, highway construction, traffic congestion, fuel and energy consumption, air pollution, noise, vehicle-oriented development, social segregation, and other problems caused by urban sprawl

Concluding Remarks
Like many cities around the world, the Tehran metropolitan region is growing rapidly.How it grows, including neighborhood density, diversity (mix) and roadway network design, can have many economic, social and environmental impacts.This study uses an integrated land use and transportation model to investigate how various land use factors affect travel activity, and therefore various transport costs.This research contributes to our understanding of urban development impacts, which has practical implications for policy and planning decisions.
This research indicates that Metropolitan Tehran's growth does not enjoy balanced and sustainable conditions.The results support the Tehran's Comprehensive Plan's goals to apply growth management in order to create more accessible and multi-modal communities in order to reduce vehicle travel and associated costs.However, consumer surveys indicate that a significant portion of households currently prefer to live in more suburban, and therefore more automobile-dependent, neighborhoods.Policies must respond to these preferences, for example, by creating more walkable and transit-oriented suburbs, and to find incentives that would encourage more middle-income households to choose accessible and multi-modal urban neighborhoods over sprawled locations.
Robust methodologies for measuring accessibility, the key concept that links land-use and transportation, are needed to adequately evaluate the effects of land-use policies on transportation and vice versa.Finally, the potential impacts of policy responses on urban location and mobility decisions, as opposed to the known impacts of individual lifestyles and preferences, and the implications for modeling techniques can be an interesting line of inquiry in future research on Tehran city.

Figure 1 .
Figure 1.Conceptual framework of relation between urban growth structure and travel behavior Source: Authors' work

Figure 2 .
Figure 2. Geographical location of Tehran and it's Characteristics from up Source: Authors' work (left); Tehran Comprehensive Transportation and Traffic Co. http://trafficstudy.tehran.ir(right)

Figure 3 .
Figure 3. Trend of population increasing in Tehran from 1926 to 2015 Data source: Statistical Center of Iran.http://www.amar.org.ir/english

Figure 4 .
Figure 4. Trend of changes in Tehran's daily trips from 2006 to 2015 Data source: Tehran Comprehensive Transportation and Traffic Co. http://trafficstudy.tehran.ir

Figure 5 .
Figure 5. Highway and Street network Figure 6: Street, Metro, Bus and BRT network Source: Tehran Comprehensive Transportation and Traffic Co. http://trafficstudy.tehran.ir

Figure 8 :
Figure 8: daily trips by modes of transportation Figure 7. Share of vehicles in traffic production Data source: Tehran Comprehensive Transportation and Traffic Co. http://trafficstudy.tehran.ir

Figure 9 .
Figure 9. Mobility condition in Tehran urban region by drawing on geographical distribution Source: Authors' work.
Figure 10.Public Values and Attitudes Travel Demand = Density + Diversity + Design + Destinations (1) • Density = Percent Change in [(Population + Employment) per hectare] • Diversity = Percent Change in {1 -[ABS(b * population -employment) / (b * population + employment)]} • where: b = regional employment / regional population • Design = Percent Change in Design Index • Design Index = 0.0195 * street network density + 1.18 * sidewalk completeness + 3.63 * route directness where: 0.0195 = coefficient applied to street network density, expressing the relative weighting of this variable relative to the other variables in the Design Index formula, street network density = length of street in meter/area of neighborhood in hectares 1.18 = coefficient applied to sidewalk completeness, expressing the relative weighting of this variable relative to the other variables in the Design Index formula, sidewalk completeness = length of sidewalk/length of public street frontage 3.63 = coefficient applied to route directness, expressing the relative weighting of this variable relative to the other variables in the Design Index formula, route directness = average airline distance to center/average road distance to center Destinations (accessibility) = Percent Change in Gravity Model denominator for study TAZs "I":Sum[Attractions(j)*Travel Impedance (i,j)] for all regional TAZs "j"

Table 1 .
Costs of Sprawl and Automobile Dependency

Table 2 .
Approaches, Theories and concepts in Land use-transport interaction modeling

Table 3 .
Correlations between automobile-dependency variables in Tehran regionsThe region's settlement pattern has represented the lifestyle preferred by most Tehran families in the past few decades.Pearson Correlation among different types of urban land uses as mixed land use index is shown in table 4. Here, significant correlation among different types of land uses in all districts means land use mix, but insignificant correlation means low mix in urban land uses.These findings indicate that Tehran's development pattern results in low mix (or described differently, large separations between different land use types, such as housing and commercial services), which increases travel distances, mobility, social seclusion and automobile dependency.

Table 4 .
Correlations between land-uses types in Tehran regions

Table 5 .
Tehran's urban growth quality base on urban indexes

Table 6 .
Basic data and Study Parameters Authors' work with base data source from Tehran Comprehensive Transportation and Traffic Co.: 2015; http://trafficstudy.tehran.ir& Statistical Center of Iran http://www.amar.org.ir/english

Table 7 .
Measuring of Land Use factors Impacts on VMT