- Email: [email protected]
The influence of integrated space–transport development strategies on air pollution in urban areas
Transportation Research Part D 44 (2016) 134–146
Contents lists available at ScienceDirect
Transportation Research Part D journal homepage: www.elsevier.com/locate/trd
The influence of integrated space–transport development strategies on air pollution in urban areas Lasmini Ambarwati a,b,⇑, Robert Verhaeghe a, Bart van Arem a, Adam J. Pel a a b
Department of Transport and Planning, Faculty of Civil Engineering and Geosciences, Delft University of Technology, Stevinweg 1, 2628 CN Delft, The Netherlands Department of Civil Engineering, Faculty of Engineering, Brawijaya University Malang, East Java, Indonesia
a r t i c l e
i n f o
Article history:
Keywords: Space–transport development Controlling urban sprawl Air pollution Improvement of public transport Job housing balance
a b s t r a c t The phenomenon of urban sprawl has strong impacts on transport performance and accessibility and causes an increase of air pollution. Effective control of urban sprawl requires an integrated approach comprising urban transport and land-use planning. Current research is insufficient to demonstrate the effects of urban sprawl on travel behavior and air pollution emission. The present paper examines the potential of an integrated approach on space–transport development strategies with the aim of increasing accessibility and reducing air pollution. A combination of space and transport strategies has been simulated for the rapidly expanding city of Surabaya. A comparative analysis of the impact of those cases indicates the promising potential alternatives to minimize the phenomenon. The transport options considered are combinations of Public Transport (PT), comprising Mass Rapid Transit (MRT), Light Rapid Transit (LRT), and Bus Rapid Transit (BRT). The options for urban structure include a compact zone development for the city, as formulated by the city planning agency, and a polycentric city set-up based on a job-housing balance aimed at minimizing the house-job distance. The results indicate that the polycentric city structure has the potential to make public transport work successfully for the city of Surabaya. This city structure creates a trip demand pattern which matches citizens’ PT preferences. Compared to the current situation, the combination of such a city structure with an expansion of PT systems would lead to a considerable improvement of transport performance, i.e. a PT mode share, a mean commute distance, and a significant reduction in emissions. Ó 2016 Elsevier Ltd. All rights reserved.
Introduction Rapid expansion of city areas is occurring world-wide. The phenomenon of urban sprawl plays an important role in the build-up of urban areas, and it refers to a complex pattern of land use, transportation patterns, and social and economic development, influencing living conditions. Urban sprawl is characterized by a rapid expansion of residential area at the outskirt of the town, causing a spatial mismatch of jobs and residential dwellings, further causing imbalance in transport with high dependence on automobile (Duncan, 1989). Banister (1996) looked at integrating activities into high density areas and
⇑ Corresponding author at: Department of Transport and Planning, Faculty of Civil Engineering and Geosciences, Delft University of Technology, Stevinweg 1, 2628 CN Delft, The Netherlands. Tel.: +31 15 27 89575. E-mail addresses: [email protected], [email protected] (L. Ambarwati), [email protected] (R. Verhaeghe), [email protected] (B. van Arem), [email protected] (A.J. Pel). http://dx.doi.org/10.1016/j.trd.2016.02.015 1361-9209/Ó 2016 Elsevier Ltd. All rights reserved.
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
135
land-use mixes in order to reduce vehicular travel. Some effects of this phenomenon are the increase of automobile dependence, air pollution, automobile accidents, and pedestrian injuries and fatalities (Frumkin, 2002). Since the beginning of the 20th century, urban sprawl has been identified in most Indonesian cities. It reflects a pattern of urban development with increasing settlement growth in the suburbs. The phenomenon of urban sprawl occurs in the city of Surabaya as case study area, are characterized by an estimated 38% of the population lives in the suburbs (Statistic Bureau of Surabaya City, 2010); most of them commute every day to work in the central urban area. This extensive growth in economic and residential development has significant consequences for mobility. Traffic congestion in the city of Surabaya has strongly increased since the late 1990s; predominantly due to increasing numbers of motorcycles and private cars. Current mode choice by residents: private car (30%), motorcycle (62%), and other vehicle types (8%, mostly minibuses). This situation results in significant increase of time, cost and productivity losses, and an increase in air pollution. Sustainable transportation linked to spatial planning is a major consideration for improving mobility. Such a strategy is expected to change modal split from private vehicles to public transport, to reduce the amount of air pollution from motorized vehicles, and to lessen travel time and distance. Private road transport is currently a substantial source of environmental pollution and traffic congestion in urban areas. Another strategy, taking into consideration urban structures, based on compactness, was expected to produce lower levels of pollutant emissions, particularly CO2 emissions, as well as fuel usage, travel distance and travel time (Marquez and Smith, 1999). They explained the influence of city structure on air quality by developing a framework consisting of various components, including a GIS database, a land use-transport-environment module and an airshed model. Stone (2008) revealed that large metropolitan regions rank higher on a quantitative index of sprawl than spatially compact metropolitan regions. This study has promoted land use strategies concentrating on more compact urban areas, intended to solve problems related to urban air quality by technological emission controls. Martins (2012) also presented a sustainable urban structure related to air quality. The impact of various urban growth patterns on air quality of the Porto urban region in Portugal has been assessed by applying the MM5-CAMx modeling system. The analysis concluded that an urban structure with sprawl has a high level of pollutants. On the other hand, a compact structure has a dense population, and increasing pollution (per space unit) due to the high concentration level. He indicated that consideration of land use strategies, such as the compact regions is designed to increase urban air quality. Regarding transport strategies, numerous studies have been conducted on transport demand management (TDM) with the aims of reducing the air pollution from motor vehicles. The effectiveness of two demand management measures, i.e. road pricing and the vehicle quota scheme (VQS), was revealed as instrumental in controlling both congestion and automobile ownership (Chin, 1996). Mitchell (2005) explained that there was a significant level of environmental inequality in Leeds, in the UK. Briefly, environmental inequality was reduced by an analysis of transport strategies with natural fleet renewal and road-user charges. The changing modal split from private vehicles to public transport and decreasing travel demand are benefits in promoting emission control technology and clean fuel. These studies on transport strategies, such as TDM and the promotion of technology to control emission, found that those strategies were insufficient to control the phenomenon of urban sprawl. Regarding travel behavior and pollution emissions, studies have focused on the impact of mode choice on energy consumption and pollution emission. Coefficients of energy consumption and emissions are influenced by the different sizes and ages of engines (Chiou et al., 2009; Chiou and Chen, 2010). These authors have designed an integrated model for cars and motorcycles, assessing choice behavior related to ownership, type and usage. They have proposed potential reductions of air pollution particularly HC and CO emissions by manipulating variables, such as an increase of ownership and user costs, and improving transit services. Nejadkoorki et al. (2008) have developed an approach for modeling CO2 emissions related to traffic-generated emissions for major roads in Norwich, in the UK. They suggested that urban restructuring, developing road networks, and changing traffic demands would reduce air pollutions. Early studies focused on the structure of vehicles and travel behavior to minimize air pollutions. They indicated the need to consider the integration of urban development and transport systems to reduce air pollutions. An integrated model to reduce air pollution by the improvement of transport technology was applied in Beijing, China. An intelligent transport management system could mitigate the proportion of pollution from vehicles and improve emission performances (Costabile and Allegrini, 2008). Loo and Chow (2011) explained that an analytical framework focusing on population patterns and job relocation policies has been applied in Hong Kong. The framework was intended to quantify potential commuting savings and environmental benefits by employing different job policies with different rates and patterns of population decentralization. Loo and Chow (2011) also discovered that modification of the urban form with three spatial job policies, and job decentralization had significant effects on shortening commuting patterns and the realization of a sustainable transportation in the city. Their previous research (Loo and Chow, 2008) demonstrated that providing good public transport should be connected with the new growth areas due to the increase of attractiveness of working in those areas. The improvement of public transport aims to increase connections in the new growth areas and to reduce their dependency on the city center. However, the above studies were not clearly addressing the importance of integrated space–transport development strategies in minimizing air pollution and contributing to sustainable urban transportation with a design of various urban structures. An insufficient number of studies have considered the integration of urban planning and transport systems as the reduction of air pollution. Therefore, this study focuses primarily on the design of integrated space–transport development strategies which will not only reduce congestion, but also have a significant impact on air emissions. In this study, an integrated
136
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
approach consists of a combination of space–transport development and impact assessment. The focus is on identifying a promising options based on an analysis of combinations of transport and urban land-use development. Emissions from all transport modes are traced. A comparison of choice behavior of residents and emission loads for the various alternatives provides some insight into the best strategy to increase mobility and to reduce air pollution. This study has been based on a survey to estimate the parameters in a four-step transport model, focusing on mode choice. Parameters of empirical analysis of heterogeneous traffic flow were employed in traffic simulations (Ambarwati et al., 2014a). The results of the model will be compared to residents’ preferences related to transport mode choice. In the face of rapid growth in the use of cars and motorcycles, it is necessary to consider the strategy of space–transport development in order to reduce the number of private vehicles and the high level of air pollution caused by private vehicles. In general, the motorcycle is a relatively fuel efficient but highly polluting mode. The results of this study are expected to inform the design of a suitable city structure linked to improvement of PT system in order to minimize the phenomenon of urban sprawl in developing countries and to assist in policy making with regard to addressing air pollution as part of the control of urban sprawl; a recommendation is made regarding the best way to minimize air pollution. This paper is organized as follows: data collection and methodology are elaborated in Section ‘‘Data and framework”. Background information on the public transport and urban spatial development of the city of Surabaya is also described in Section ‘‘Data and framework”. Section ‘‘The results of simulation analysis” details a simulation analysis of transport performance for a combination of urban structure scenarios and improved public transport systems. Section ‘‘Assessment of emission load for each case” discusses the estimation of emission parameters and compares the impacts of the different spatial structure and PT options. Section ‘‘Conclusions and observations” presents conclusions and recommendations for further analysis. Data and framework This section describes the data collection for the study area and the methodology which uses a combination of two models, i.e. the Java Spatial Model (JSM) and OmniTRANS. This section consists of data needed to estimate emissions, the parameters of the four-step transport model for each scenario, and the background for the set-up of the transport and spatial developments. Data The case study area is the city of Surabaya (Fig. 1), the capital city of the province of East Java, comprising 31 districts and 163 sub-districts (called ‘‘desa”) with a total area of 327 km2. The city has a population of approximately 3 million, a high density of more than 11,000 persons per km2, it can be considered a highly urbanized area (Statistic Bureau of Surabaya City, 2010).
Fig. 1. Location of study area (the city of Surabaya) in the province of East Java, Indonesia.
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
137
Build-up areas have concentrated in the current larger urban centers in most big cities in Indonesia since the 1980s, such as Jakarta, Bandung, Semarang, Yogyakarta, Medan, and Makasar, which have similar conditions and urban transport problems with the city of Surabaya. Currently, congestion and air quality have become major problems in these urban centers; this will grow out of proportion when these urban centers approximately double in size in the next twenty years. This rapid and uncontrolled growth has strong consequences for mobility for the residents of these metropolitan areas. The development of transport infrastructure is lagging behind the times, with resulting large scale congestion and air pollution. In the future, the urban development of most big cities in Indonesia should be managed a sustainable growth by controlling land use integrated with a transport network in order to decrease huge congestion and pollution, and to maximize accessibility. The spatial unit for the models is the desa, so data is collected at the desa level. The transport and spatial models require socio-economic and transport behavior data as well as associated schematizations. The 2010 Census data have been used to analyze population, employment, attractiveness variables and land-use for each of the 163 desa in the city. Transport data was collected by means of an extensive survey (2012) and the video recording of traffic. From the questionnaires distributed, a total of 554 respondents, a representative of residents of 163 desas. Analysis was based on data collected from all respondents. The questions were divided into three groups: socio-economic background, trip characteristic and transport mode choice, residents’ preferences regarding public transport and living conditions. Planning methodology The conceptual framework, presented in Fig. 2, describes the computational steps to quantify the components of integrated plans for addressing the urban sprawl phenomenon. It consists of the application of two models to simulate the performance of alternative combinations of spatial-transport development. In the present paper, the output focuses on the emissions for those combinations. The models employed in this study are OmniTRANS, to set up the transport model and to assess the environmental impact, and the Java Spatial Model (JSM), to simulate spatial settlement. The JSM was calibrated in previous research (Verhaeghe et al., 2009), while the calibration of the transport model was discussed in previous research (Ambarwati et al., 2014b). A set of simulations, comprising a spatial settlement strategy and a transport system set-up were carried out. The impacts from this set of simulation cases were compared and formed as input for the formulation of strategic options for integrated development of the city. Fig. 2 presents an overview of this methodology. A total of four spatial settlement scenarios (scenarios for the transport system) and five alternative (public) transport situations are combined into nine simulation alternatives (cases). The cases are Spatial scenarios: (1) Existing (2010) settlement, (2) Settlement 2030 based on a continuation of current trends, (3) Settlement 2030 with the consideration of a socalled compact zone in the western part of the city, and (4) Settlement 2030 with a polycentric city. Transport alternatives: (a) Existing situation (2010), (b) Situation 2030 with currently planned (current trend) transport options, (c) Situation 2030 currently planned and BRT, (d) Situation 2030 currently planned + MRT + LRT, and (e) Situation 2030 currently planned + all PT systems (BRT, MRT, LRT) Nine cases are simulated, i.e. 1a, 2b, 2c, 2d, 2e, 3b, 3e, 4b, 4e. The computations in the 3 modules can be summarized as follows: Firstly, JSM (Java Spatial Model): this model used the input of the statistics for population and employment from the 2010 census, and applied socio-economic trends and spatial planning interventions. The output of JSM (projection of population, employment, and housing area) was used to generate
Fig. 2. The conceptual framework of space–transport development.
138
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
the traffic demand which was the input for the transport model of each scenario. Secondly, OmniTRANS transport model: the input for OmniTRANS consisted of zonal data (desa), transport networks and user behavior. The zonal data was obtained from JSM and the network was set up with a GIS application. The user behavior consisted of: trip production and attraction parameters, distribution functions, BPR-functions, and VOT (value of time). The parameters of the different functions were designed in a previous study (Ambarwati et al., 2014b). In the simulation of the OmniTRANS model, residential density was inputted to determine the number of trips generated from the residential areas in each zone. The various transport modes consisted of cars, motorcycles, public transport (bus, minibus, MRT, LRT, and BRT), bicycles, and walking. Walking was considered as access and egress mode to public transport system. Thirdly, impact assessment module: computation of emissions based on the transport flows simulated with OmniTRANS; in particularly the load of emissions such as CO, CO2 , NOx, pm10, and HC. This study only focused on CO emissions. Spatial development scenarios Simulation of spatial development strategies is based on the existing spatial panning and transport network, the design of a compact zone in the particular regions related to the Spatial Masterplan of Surabaya, and design of a polycentric structure intended to shorten the travel distance. Four alternative city structures are considered with distinct spatial planning. They are elaborated below. Current situation and current trend The current situation 2010 describes the current settlement in for the city of Surabaya. The spatial schematization comprises 163 administrative zones (desas) of the city of Surabaya; 5 regions are considered (central, north, south, west, and east) as shown in Fig. 3. The total area of the city is approximately 33,637 hectares, about 65% of which can be found in the west and east regions. The built-up area in these regions is less than 15%, meaning that there is more potential development in these regions than in other regions of the city. Since the 1980s, settlement development has extended to the outskirts of the urban area, particularly on the western and eastern sides of the city. There are 135,000 households living in the suburbs. Some zones in the western regions have the highest job density due to new industrial enterprises. These regions create a significant number of jobs for the city, approximately 20% of total employment. They have a low population density resulting in a high employment-population ratio. The transport infrastructures in these regions are inadequate, and as a consequence it takes more than 90 min by motorized vehicles and more than 110 min by PT to reach the city center. The current trend 2030 comprises the projected development of Surabaya City in 2030, based on a continuation of the current situation and implementation of current planned transport interventions; further an annual growth of 5% GDP
Fig. 3. ID for the desa level and regions of the city of Surabaya.
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
139
and 1% employment has been adopted over a 20 year period (2010–2030). In this scenario, the population will reach approximately 4.5 million residents, with the highest population in the western regions. Employment growth is more equally divided over the city with a total number of almost 1.6 million jobs. The current trend 2030 includes the expansion of the transport network with the extension of eastern ring road, a new collector, and local roads in the western regions. This is expected to increase residents’ ability to access to the city center from the eastern and western regions. Compact zone strategy – balanced distribution of settlement Creating a compact zone, aiming at a high density with mixed land-uses and complementary functions such as housing, shopping and offices, is considered a sound strategy to reduce energy use in households and the extent of everyday travel in densely populated area. The central urban areas activities represented the highest level of leisure-time travel by plane (Holden and Norland, 2005). Compact zone strategy is considered in the Spatial Masterplan of Surabaya. Observation on the distribution of employment and residential density in the city indicates relatively high employment in the western part of the city attracting workers from other parts of the city. In this part of the city, the local government has set up a new factory centers, and this has led to a higher employment density in this area of the city, as seen in Fig. 4a. Fig. 4b illustrates how this part has a high employment-population ratio due to a low population density, yet maintains a high employment density. Currently, the residential areas is located the certain areas, away from industrial areas and other business areas. Under the idea of compact growth and to better manage the balance between employment and residential density, the compact zone strategy proposes the extensive development of residences, shopping facilities, and educational centers only in the western part. It means that a new sub-center in the western part would be designed to increase the number of workers living in this part and the importance of intra-zonal commuting movements. Currently, this part has less accessibility than other parts, The expectation is to reduce travel length and the number of trips to the city center from this part, as a result the effect of urban sprawl can be minimized and residences’ accessibilities can be encouraged. Such a spatial strategy is adopted in the present analysis of urban development options for the city and the consequences on transport has been quantified. The compact zone strategy in the present analysis has been designed using a similar overall current trend 2030 with GDP growth 5% and employment rate of 1%. However for this zone (the western part of the city) the ratio of employment to population increases 3% from the current trend 2030 (Ambarwati et al., 2014b). Job-housing balance (JHB) – polycentric city Reducing work trip distance and an increased use of public transport are primary goals for planners ad policy makers (Horner, 2004; Black, 2010) towards a sustainable (economic, environmental, and social) urban development. Various researchers (Ewing and Cervero, 2001; Cervero, 2002; Cervero and Duncan, 2006; Guo and Chen, 2007; Levine et al., 2012) argue that the job-housing perspective should aim to reduce the spatial disparity between homes and jobs, so raising the potentially beneficial effect on sustainability, i.e. shorter commutes and reduction in the use of private vehicles. Breheny et al. (1992) emphasized that dwellings, workplaces, public and private services should be developed around public transport nodes intended to reduce commuter distance. A design of job housing balance (JHB) is proposed in this study aimed to promote job location close to housing developments. This study plans a polycentric structure where both population and jobs are decentralized by designing 4 sub-centers in all parts of the city and associating these parts with the expansion of local PT (bus and minibus) as feeder systems. Under this urban form, intra-zonal commuting movements are expected to emerge as a new travel pattern in all parts of the city. The locations of the current main center and the 4 sub-centers are indicated in Fig. 5. The new sub-centers are designed for mixed land use, aimed at reducing residents’ dependence on the city center for their subsistence activities (working, shopping, schooling). Development of poly-centers has been carried out in order to shorten the distance between job and housing location, involving the design of new sub-centers in the different regions of the city. This strategy is also expected to reduce travel length for each region to the city center and to increase travel to the nearest sub-center. As a result, it is necessary to encourage residents to increase their use of walking, bicycling, and transit usage due to land use mix and closer travel within smaller spatial areas (new location patterns with a new sub center in each region of the city). A minimum commute concept (called MIN: minimum average commute in length) has been used to characterize a particular urban structure. It represents a jobs-housing benchmark which has its origin in urban economic theory. MIN represents a theoretical work travel pattern in which workers are assigned to alternate workplace locations, resulting in work trips with the lowest possible length. Urban sprawl disturbs the job housing balance and increases the MIN. The MIN can be quantified using a linear optimization called the transportation problem (Hitchcock, 1941; White, 1988; Giuliano and Small, 1993). The analytic framework can be formulated as follows: Minimize MIN
MINv ¼
n X m 1X cij xij W i¼1 j¼1
ð1Þ
The objective function minimizes the average travel length in the assignment of workers (xij ) between homes and jobs. Subject to
140
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
Fig. 4. Job density (a) and population density (b) in each zone of the city. m X xij ¼ Oi ;
8i ¼ 1; . . . ; n
ð2Þ
8j ¼ 1; . . . ; m
ð3Þ
j¼1 n X xij ¼ Dj ; i¼1
xij P 0 X bijv cijv cij ¼ v
ð4Þ ð5Þ
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
141
Fig. 5. Location of main center (central business districts (CBD)) and new sub-centers.
where n = the number of origin or home zones, m = the number of destination or employment zones, W = the number of workers in the study area, Oi = the total number of workers living in zone i; Dj = the total number of jobs in zone j; cij = the travel length between home site i and job site j; xij = the number of workers living in zone i and working in zone j; bij = the choice probability of mode v home site i and job site j. At the opposite spectrum of the MIN, a theoretical random commute can be considered. The assignment of workers between homes and jobs is random with an equal probability that any home will be matched to any job. To explore the ultimate potential effect of urban form for Surabaya, a MIN has been prepared associated with a polycentric city structure in which 4 sub-centers are added to the existing main center. In this development scheme, the average commute will be shortened because of the high accessibility to new jobs located in the new sub-centers and will encourage a strong expansion of local PT (minibuses) as feeder systems. The design of a polycentric urban structure with 4 sub-centers is expected to become a model which can be used by other big cities in Indonesia. Improvement of public transport (PT) systems Several needed improvements of public transport have been identified. The set up and modeling of those public transport systems, including model calibration by means of an extensive survey, have been elaborated in previous study (Ambarwati et al., 2014b). Currently, the Surabaya public transport network consists of minibusses (paratransits) and busses. There are 68 available minibus routes, and 22 bus routes (minibus accommodates 8–12 passengers, while bus has 50–55 seats). The frequency of minibusses and busses are approximately 20–25 veh/h and 5–6 veh/h respectively. The current performance of public transport such as capacity, quality, and efficiency was measured by conducting an on-board survey for bus and minibus/paratransit. The results of this survey were used as input for the design of improvements to public transport by employing the OmniTRANS model. Each part of the city is served by approximately 26 minibus routes, except the southern part of the city, which is accommodated by 10 minibus routes. The recommended improvements comprise a MRT (monorail), a LRT (tram line), and a BRT (Bus Rapid Transit) with a grid structure. For analysis, the different systems have been combined into 3 cases (Section ‘‘Planning methodology”): BRT, MRT + LRT, and all PT systems. The current local PT is associated with the improvement of the PT system as a feeder system. Fig. 6 presents the schematizations for the 3 cases. A BRT (Bus Rapid Transit) system was planned on the main roads of the city with the proposed routes (see in Fig. 6a). The BRT system should be more integrated with the road network of pedestrians to make the system more attractive to them (Fitts and Midgely, 2010). The properties of this BRT system were proposed as follows: 80 seats, 12 vehicles per hour per route, with a ticket price of about $0.51, an average speed 30 km/h, and stop spacing of 0.5–2 km. Due to the current amount of congestion occurring in Surabaya, the city of Surabaya is planning a Mass Rapid Transit (MRT) and tramline (LRT), as illustrated in Fig. 6b. The aim of the MRT is to provide a sustainable transportation system in Surabaya, to reduce the congestion and to provide alternative options of public transport that are safe, comfortable and scheduled (Department of Transport of Surabaya City, 2013). The properties of the proposed MRT and LRT are: a ticket price of approximately $0.82, stop spacing of 0.5–2 km, seat capacity of 177 seats for the MRT and 100 seats for the LRT, frequency of 12 veh/h, and an average speed of 55 km/h for the MRT and 35 km/h for the LRT.
142
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
Fig. 6. BRT-system with grid structure (a), development of MRT and tram lines (b), and all PT-systems consisting of BRT, LRT, MRT and feeder systems in the future for Surabaya City (c).
143
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
Based on integration of the previous ideas, improvement of all PT systems consists of MRT (monorail), LRT (tram line), BRT (Bus Rapid Transit) and buses and minibuses as feeder systems (see in Fig. 6c). The improvements to public transport are expected to encourage the users of cars and motorcycles to shift their transport mode to the public transport systems. In Section ‘‘Planning methodology” nine system cases have been defined comprising a combination of a spatial structure (at a particular time: present or 2030) and a public transport option. Each of these cases has been simulated. The simulated flow patterns for the different modes of transport are used to compare the significance of the improvement of various PT systems linked to changes of land use patterns. The comparison involves assessing the performance of the different cases as well as the emission loads associated with these transport patterns. This comparison is made in order to reveal the extent of the change of spatial or transport system development intended to increase residents’ mobility and to reduce air pollution. Section ‘‘The results of simulation analysis” discusses the transport performance. Sections ‘‘Assessment of emission load for each case” and ‘‘Conclusions and observations” analyse the emission loads.
The results of simulation analysis Important features of the transport system, concerning accessibility and air quality are modal split and commuting distance. Alternatives which promote a modal shift towards public transport and decrease commuting distance will positively contribute to increased accessibility and air quality. Below, Section ‘‘Modal split” discusses the effects of the nine simulation cases on modal split, and Section ‘‘Commuting distance” presents the results on commuting distance.
Modal split Improvements in public transportation systems are intended to increase the people’s use of public transport in each case. Each case here examines the change of modal split for each vehicle class. Table 1 presents the results of modal split for these different cases. The following points can be noted from the chart: – An improvement of PT has only a slight effect on the mode share of PT (C2 versus C5); it reduces slightly the use of private cars and the share of motorcycles. – The compact zone spatial intervention has on its own little effect on the mode split for PT (C2 versus C6). The alternative design of a compact zone with an added improvement of PT encourages this mode split (C2 versus C7); it can be observed that C7 has practically no effect on the motorcycle mode, which remains by far the dominant mode. – A drastic change in urban structure (job housing balance (JHB)), C9 has a drastic effect on the mode split for PT: the PT mode share increases to 70%. The alternative is associated with a strong increase in the bicycle mode. The mode shares of motorcycles and private cars reduce drastically. In particular the strong reduction in motorcycle use is remarkable because it remains dominant in all other cases. An extensive reduction of use of the motorcycles in some zones, such as zones in the south, west and east sides of the city occurs because of accommodated by MRT and BRT lines. This can be explained by the strong improvement of home-job trip situation and the attractiveness of the new centers: the demand for trips in this urban structure matches very well with residents’ preferences of use of public transport in Surabaya. – Case C9 also corresponds to the preference of residents for a short trip distance, as communicated through the survey. Forty percent of respondents respond that they prefer to use PT for trips of 8.5 km or less. This means that change modal split with enlightened spatial development, i.e. polycentric city planning with designing 4 new sub-centers in the city increase in users of PT. In short, Case C9 demonstrates the importance of urban structure on accessibility and obviously represents part of an ideal solution for development of the urban structure of Surabaya. Provision of sufficient PT is necessary to take over from other modes but an adequate urban structure is needed to make a new PT system successful.
Table 1 Modal split for each scenario (%). Transport mode
PT Motorcycle Car Bicycle 1
Spatial development scenarios
Transport options
Cases (C1) current base
(C2) current trend
(C6) compact zone
(C8) JHB1
(C3) BRT
(C4) MRT and LRT
(C5) all PT
(C7) all PT + compact zone
(C9) JHB + all PT
5.82 58.73 34.88 0.57
5.62 63.39 30.23 0.75
5.66 64.3 29.2 0.82
62.48 23.3 13.02 1.23
5.69 60 28.86 5.48
7.72 58.7 28.7 5
8.51 59.8 26.58 5.04
9.23 60.3 25.39 5.09
70.66 10.35 6.43 12.57
Includes a strong expansion of localized PT (small buses).
144
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
The design of the city structure is aimed to reduce travel to the city center and to minimize residents’ dependence on the city center, as explained in the next section.
Commuting distance In this study, we assessed the minimum average commute (MIN). The values were required to evaluate the significance of the space–transport development strategies in shortening the trips. Table 2 presents the average commuting distance for the different cases. The following observations can be made: – In Cases C1 and C2, a high number of commutes take place to the city center, particularly by motorcycle. – The case with improvement of public transport (C5) has had a significant effect on reducing the average commuting distance of cars in contrast to motorcycles, meaning that what used to be long car trips could instead be trips making use of PT. – The compact zone case C7 has a considerable effect on shortening the commuting distance. – The polycentric urban structure and improved PT systems, seen in case C9, results in the lowest commuting distance for all transport modes; the commute distances are about half of the present (C1) and future current trend (C2) cases. It is clear that case C9, with a very high PT mode share and short travel distance, will have a strong effect on residents’ mobility. This means that C9 is expected to have an effect on the emission load. The emission load is elaborated next in S ection ‘‘Assessment of emission load for each case”.
Assessment of emission load for each case Unit emission loads (g/veh km) are vary for different modes and different speeds. Relevant emission substances are CO, CO2 , NOx, pm10, and HC. An analysis of the air quality of the city of Surabaya indicates that CO emission is dominating the air pollution problem because of the extensive use of motorized vehicles. Most motorcycles emit CO, NOx and HC, while cars exhaust 30–50% CO and HC (Badami, 2005). Chiou and Chen (2010) also revealed that HC and CO emission levels are primarily determined by car and motorcycle exhaust. Therefore the analysis of this research will focus on CO emissions. The weighted emission values for each vehicle type depend on the average standard speed of each vehicle class for each road type (motorway, urban, and rural). The values were analyzed using the procedure for determining air emission rate for each vehicle class at the micro level as stipulated in the guideline of The Ministry of Public Work (1999). The emission rate ((g/veh km) is expressed as
qCO ¼ 867:92 U 0:8648 v
ð6Þ
where U v is average speed standard for each vehicle class for each road type in km/h. The emissions for a particular traffic flow over a particular stretch of road by a particular vehicle class are determined by
Air pollutant ðgÞ ¼ traffic flow of each vehicle class ðvehÞ emission value for each parameter for each type of vehicles ðg=veh kmÞ road length km
ð7Þ
The CO emission load has been computed by simulation of the OmniTRANS model, based on the simulated flows on the transport network, for various cases; the results are presented in Table 3. The following can be noted:
Table 2 Average commute for each alternative (in km). Average commute
PT Motorcycle Car 2
Spatial development scenarios
Transport options
Cases (C1) current base
(C2) current trend
(C6) compact zone
(C8) JHB2
(C3) BRT
(C4) MRT and LRT
(C5) all PT
(C7) all PT + compact zone
(C9) JHB + all PT
4.08 5.52 4.95
4.8 5.78 5.52
3.44 3.94 3.57
2.28 3.05 2.6
3.83 6.17 4.34
4.5 5.33 3.49
3.65 5.43 3.46
2.89 3.44 2.89
2 2.57 2.22
Includes a strong expansion of localized PT (small buses).
145
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146 Table 3 The average CO emission for all cases (in tons/peak-hour). Transport mode
2010
2030
Spatial scenarios
Car Motorcycle PT Total
PT system options
(C1)
(C2)
(C6)
(C8)
(C5)
(C7)
(C9)
22.8 53.4 84.3 160.5
65.2 124 45.1 234.4
70.2 118.9 22.7 211.8
1.4 4.8 10.9 17.2
19.8 43.9 19 82.7
14.2 29.3 8.5 52.8
0.4 0.8 9.6 10.9
– For the present (C1), the estimated peak hour load of CO from cars, motorcycles and PT for the whole city is around 160 tons. This is expected to increase dramatically in the future (C2) to 234 tons if no spatial and PT interventions are implemented. This will be primarily caused by an increase in the use of private vehicles. These findings are confirmed by Cervero (2000), who explained that exposure levels and thus health risks are lower, on the other hand tailpipe emissions and fossil-fuel consumption will continue to increase. – Improvement of the PT system (C5) has little effect on decreasing the extents of load of private vehicles (car and motorcycle), as reflected in a slight reduction of CO emissions. – The design of a compact zone (C6) does not provide a significant reduction in emission load. – The hypothetical polycentric city structure and expansion of PT systems (C9) results in a drastic change in the emission load (90% reduction), due to the combined effect of short commuting distances and the very strong transition to public transport. In comparative analysis of all alternatives, C9 yielded, as expected, the lowest level of air pollutant concentrations (CO emission), average commute, and the highest number of passengers using public transport system. Because of high housing development in the suburbs, the design of new sub-center in each region should be undertaken and connected to the city center by public transport systems. The real obstacles to achieving that sort of development pattern are cost, coordination between the authorities for housing and job developments. Realization of the development pattern should be supported by regulation of spatial development and management of land use. The regulation should be recommended to the local government of Surabaya.
Conclusions and observations The results of the analysis indicate that an integrated approach, comprised of interventions in urban structure and public transport, is necessary to address the mobility and emission problems in fast growing urban areas and the effects of urban sprawl. It does not seem that an improvement of public transport would substantially reduce the use of motorcycles, in which residents currently see as the best way to deal with congestion. It therefore at present becomes the dominant transport mode (59.8%) and is projected to remain so until 2030. It is estimated that the compact zone approach in the current City Masterplan will have a modest impact on PT mode share: from 6% to 9%. This situation will result in a significantly shortened commute distance down to 38%, and a substantial reduction in emission load, down to 60%. Testing a polycentric urban structure for the city of Surabaya with 4 sub-centers, an associated minimized commuting distance, and an improvement of public transport, has suggested that the mode share of public transport could be increased from 6% to 70%, the commuting distance could be shortened by 53% and the emission load could be reduced by 90%. The polycentric structure shortens trips and makes public transport much more attractive to users, resulting in a travel pattern matching users’ preferences. The test with the polycentric city concept indicates that considerable improvements in transport performance and air quality are potentially possible by changing the whole land use in the city. The proposed polycentric structure and associated minimization of commute distance has the potential to decrease the motorcycle mode share to 10%. Improvement of public transport is generally considered to be the main tool for improving accessibility and reducing the emission load. On the basis of the present analysis, it can be concluded that public transport will only be really successful if associated with a suitable urban structure. This means that the key to minimizing urban sprawl is an improvement of the public transport system linked to a change in the city’s urban structure. Integration of spatial-transport strategies as well as an environmental assessment would provide long-term advantages by designing improvements of public transport systems related to settlement development. If carried forward, the integrated approach should also be based on residents’ preferences regarding the way in which these systems are developed and implemented. Further study should assess the feasibility of changing the direction of growth in the city, particularly with a consideration of residential location along transit corridors, and the influence of urban structure on transit capacity due to the change of modal split for PT.
146
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
References Ambarwati, L., Pel, A.J., Verhaeghe, R., van Arem, B., 2014a. Empirical analysis of heterogeneous traffic flow and calibration of porous flow model. Transp. Res. Part C: Emerg. Technol. 48, 418–436. Ambarwati, L., Verhaeghe, R., Pel, A., van Arem, B., 2014b. Investigating the effects of improving public transport system linkage to spatial strategy on controlling urban sprawl: evidence from Surabaya city, Indonesia. WIT Trans. Built Environ. Badami, M.G., 2005. Transport and urban air pollution in India. Environ. Manage. 36, 195–204. Banister, D., 1996. Energy, quality of life and the environment: the role of transport. Transp. Rev. 16, 23–35. Black, W.R., 2010. Sustainable Transport: Problems and Solutions. Guilford. Breheny, M.J., Association, R.S., et al, 1992. Sustainable Development and Urban Form, vol. 2. Pion, London. Cervero, R., 2000. Transport and Land Use: Key Issues in Metropolitan Planning and Smart Growth. University of California Transportation Center. Cervero, R., 2002. Built environments and mode choice: toward a normative framework. Transp. Res. Part D: Transp. Environ. 7, 265–284. Cervero, R., Duncan, M., 2006. Which reduces vehicle travel more: jobs-housing balance or retail-housing mixing? J. Am. Plan. Assoc. 72, 475–490. Chin, A.T., 1996. Containing air pollution and traffic congestion: transport policy and the environment in Singapore. Atmos. Environ. 30, 787–801. Chiou, Y.C., Chen, T.C., 2010. Direct and indirect factors affecting emissions of cars and motorcycles in Taiwan. Transportmetrica 6, 215–233. Chiou, Y.C., Wen, C.H., Tsai, S.H., Wang, W.Y., 2009. Integrated modeling of car/motorcycle ownership, type and usage for estimating energy consumption and emissions. Transp. Res. Part A: Policy Pract. 43, 665–684. Costabile, F., Allegrini, I., 2008. A new approach to link transport emissions and air quality: an intelligent transport system based on the control of traffic air pollution. Environ. Model. Softw. 23, 258–267. Department of Transport of Surabaya City, 2013. Surabaya MRT: Surabaya Mass Rapid Transportation (SMART). Technical Report. Surabaya, Indonesia. Duncan, J., 1989. The Search for Efficient Urban Growth Patterns: A Study of the Fiscal Impacts of Development in Florida. Publisher not identified. Ewing, R., Cervero, R., 2001. Travel and the built environment: a synthesis. Transp. Res. Rec.: J. Transp. Res. Board 1780, 87–114. Fitts, A., Midgely, P., 2010. An Overview of Urban Mobility and the Implementation of an Area Traffic Control System in Surabaya. Frumkin, H., 2002. Urban Sprawl and Public Health. Public Health Reports 117, 201. Giuliano, G., Small, K.A., 1993. Is the journey to work explained by urban structure? Urban Stud. 30, 1485–1500. Guo, J.Y., Chen, C., 2007. The built environment and travel behavior: making the connection. Transportation 34, 529–533. Hitchcock, F.L., 1941. The distribution of a product from several sources to numerous localities. J. Math. Phys 20, 224–230. Holden, E., Norland, I.T., 2005. Three challenges for the compact city as a sustainable urban form: household consumption of energy and transport in eight residential areas in the greater Oslo region. Urban Stud. 42, 2145–2166. Horner, M.W., 2004. Spatial dimensions of urban commuting: a review of major issues and their implications for future geographic research. Prof. Geogr. 56, 160–173. Levine, J., Grengs, J., Shen, Q., Shen, Q., 2012. Does accessibility require density or speed? A comparison of fast versus close in getting where you want to go in us metropolitan regions. J. Am. Plan. Assoc. 78, 157–172. Loo, B., Chow, S., 2008. A framework to study urban spatial structure and sustainable transport: minimum commuting, excess commuting and the potentials of greater transport sustainability. In: A seminar Paper Presented at the Hong Kong Geography Day 2008, Hong Kong. Loo, B.P., Chow, A.S., 2011. Jobs-housing balance in an era of population decentralization: an analytical framework and a case study. J. Transp. Geogr. 19, 552–562. Marquez, L.O., Smith, N.C., 1999. A framework for linking urban form and air quality. Environ. Model. Softw. 14, 541–548. Martins, H., 2012. Urban compaction or dispersion? An air quality modelling study. Atmos. Environ. 54, 60–72. Mitchell, G., 2005. Forecasting environmental equity: air quality responses to road user charging in leeds, uk. J. Environ. Manage. 77, 212–226. Nejadkoorki, F., Nicholson, K., Lake, I., Davies, T., 2008. An approach for modelling CO2 emissions from road traffic in urban areas. Sci. Total Environ. 406, 269–278. Statistic Bureau of Surabaya City, 2010. Surabaya City in Figures. Technical Report. Surabaya, Indonesia. Stone, B., 2008. Urban sprawl and air quality in large US cities. J. Environ. Manage. 86, 688–698. The Ministry of Public Work, 1999. Procedure for Prediction of Air Pollution in Micro scale from Traffic. Technical Report. Jakarta, Indonesia. Law no.017/T/ BM/1999. Verhaeghe, R.J., Zondag, B., Grashoff, P.S., 2009. Space and Water: Development of the Java Spatial Model. Technical Report. Minister of Economic Affairs. White, M.J., 1988. Urban commuting journeys are not wasteful. J. Polit. Econ., 1097–1110
Contents lists available at ScienceDirect
Transportation Research Part D journal homepage: www.elsevier.com/locate/trd
The influence of integrated space–transport development strategies on air pollution in urban areas Lasmini Ambarwati a,b,⇑, Robert Verhaeghe a, Bart van Arem a, Adam J. Pel a a b
Department of Transport and Planning, Faculty of Civil Engineering and Geosciences, Delft University of Technology, Stevinweg 1, 2628 CN Delft, The Netherlands Department of Civil Engineering, Faculty of Engineering, Brawijaya University Malang, East Java, Indonesia
a r t i c l e
i n f o
Article history:
Keywords: Space–transport development Controlling urban sprawl Air pollution Improvement of public transport Job housing balance
a b s t r a c t The phenomenon of urban sprawl has strong impacts on transport performance and accessibility and causes an increase of air pollution. Effective control of urban sprawl requires an integrated approach comprising urban transport and land-use planning. Current research is insufficient to demonstrate the effects of urban sprawl on travel behavior and air pollution emission. The present paper examines the potential of an integrated approach on space–transport development strategies with the aim of increasing accessibility and reducing air pollution. A combination of space and transport strategies has been simulated for the rapidly expanding city of Surabaya. A comparative analysis of the impact of those cases indicates the promising potential alternatives to minimize the phenomenon. The transport options considered are combinations of Public Transport (PT), comprising Mass Rapid Transit (MRT), Light Rapid Transit (LRT), and Bus Rapid Transit (BRT). The options for urban structure include a compact zone development for the city, as formulated by the city planning agency, and a polycentric city set-up based on a job-housing balance aimed at minimizing the house-job distance. The results indicate that the polycentric city structure has the potential to make public transport work successfully for the city of Surabaya. This city structure creates a trip demand pattern which matches citizens’ PT preferences. Compared to the current situation, the combination of such a city structure with an expansion of PT systems would lead to a considerable improvement of transport performance, i.e. a PT mode share, a mean commute distance, and a significant reduction in emissions. Ó 2016 Elsevier Ltd. All rights reserved.
Introduction Rapid expansion of city areas is occurring world-wide. The phenomenon of urban sprawl plays an important role in the build-up of urban areas, and it refers to a complex pattern of land use, transportation patterns, and social and economic development, influencing living conditions. Urban sprawl is characterized by a rapid expansion of residential area at the outskirt of the town, causing a spatial mismatch of jobs and residential dwellings, further causing imbalance in transport with high dependence on automobile (Duncan, 1989). Banister (1996) looked at integrating activities into high density areas and
⇑ Corresponding author at: Department of Transport and Planning, Faculty of Civil Engineering and Geosciences, Delft University of Technology, Stevinweg 1, 2628 CN Delft, The Netherlands. Tel.: +31 15 27 89575. E-mail addresses: [email protected], [email protected] (L. Ambarwati), [email protected] (R. Verhaeghe), [email protected] (B. van Arem), [email protected] (A.J. Pel). http://dx.doi.org/10.1016/j.trd.2016.02.015 1361-9209/Ó 2016 Elsevier Ltd. All rights reserved.
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
135
land-use mixes in order to reduce vehicular travel. Some effects of this phenomenon are the increase of automobile dependence, air pollution, automobile accidents, and pedestrian injuries and fatalities (Frumkin, 2002). Since the beginning of the 20th century, urban sprawl has been identified in most Indonesian cities. It reflects a pattern of urban development with increasing settlement growth in the suburbs. The phenomenon of urban sprawl occurs in the city of Surabaya as case study area, are characterized by an estimated 38% of the population lives in the suburbs (Statistic Bureau of Surabaya City, 2010); most of them commute every day to work in the central urban area. This extensive growth in economic and residential development has significant consequences for mobility. Traffic congestion in the city of Surabaya has strongly increased since the late 1990s; predominantly due to increasing numbers of motorcycles and private cars. Current mode choice by residents: private car (30%), motorcycle (62%), and other vehicle types (8%, mostly minibuses). This situation results in significant increase of time, cost and productivity losses, and an increase in air pollution. Sustainable transportation linked to spatial planning is a major consideration for improving mobility. Such a strategy is expected to change modal split from private vehicles to public transport, to reduce the amount of air pollution from motorized vehicles, and to lessen travel time and distance. Private road transport is currently a substantial source of environmental pollution and traffic congestion in urban areas. Another strategy, taking into consideration urban structures, based on compactness, was expected to produce lower levels of pollutant emissions, particularly CO2 emissions, as well as fuel usage, travel distance and travel time (Marquez and Smith, 1999). They explained the influence of city structure on air quality by developing a framework consisting of various components, including a GIS database, a land use-transport-environment module and an airshed model. Stone (2008) revealed that large metropolitan regions rank higher on a quantitative index of sprawl than spatially compact metropolitan regions. This study has promoted land use strategies concentrating on more compact urban areas, intended to solve problems related to urban air quality by technological emission controls. Martins (2012) also presented a sustainable urban structure related to air quality. The impact of various urban growth patterns on air quality of the Porto urban region in Portugal has been assessed by applying the MM5-CAMx modeling system. The analysis concluded that an urban structure with sprawl has a high level of pollutants. On the other hand, a compact structure has a dense population, and increasing pollution (per space unit) due to the high concentration level. He indicated that consideration of land use strategies, such as the compact regions is designed to increase urban air quality. Regarding transport strategies, numerous studies have been conducted on transport demand management (TDM) with the aims of reducing the air pollution from motor vehicles. The effectiveness of two demand management measures, i.e. road pricing and the vehicle quota scheme (VQS), was revealed as instrumental in controlling both congestion and automobile ownership (Chin, 1996). Mitchell (2005) explained that there was a significant level of environmental inequality in Leeds, in the UK. Briefly, environmental inequality was reduced by an analysis of transport strategies with natural fleet renewal and road-user charges. The changing modal split from private vehicles to public transport and decreasing travel demand are benefits in promoting emission control technology and clean fuel. These studies on transport strategies, such as TDM and the promotion of technology to control emission, found that those strategies were insufficient to control the phenomenon of urban sprawl. Regarding travel behavior and pollution emissions, studies have focused on the impact of mode choice on energy consumption and pollution emission. Coefficients of energy consumption and emissions are influenced by the different sizes and ages of engines (Chiou et al., 2009; Chiou and Chen, 2010). These authors have designed an integrated model for cars and motorcycles, assessing choice behavior related to ownership, type and usage. They have proposed potential reductions of air pollution particularly HC and CO emissions by manipulating variables, such as an increase of ownership and user costs, and improving transit services. Nejadkoorki et al. (2008) have developed an approach for modeling CO2 emissions related to traffic-generated emissions for major roads in Norwich, in the UK. They suggested that urban restructuring, developing road networks, and changing traffic demands would reduce air pollutions. Early studies focused on the structure of vehicles and travel behavior to minimize air pollutions. They indicated the need to consider the integration of urban development and transport systems to reduce air pollutions. An integrated model to reduce air pollution by the improvement of transport technology was applied in Beijing, China. An intelligent transport management system could mitigate the proportion of pollution from vehicles and improve emission performances (Costabile and Allegrini, 2008). Loo and Chow (2011) explained that an analytical framework focusing on population patterns and job relocation policies has been applied in Hong Kong. The framework was intended to quantify potential commuting savings and environmental benefits by employing different job policies with different rates and patterns of population decentralization. Loo and Chow (2011) also discovered that modification of the urban form with three spatial job policies, and job decentralization had significant effects on shortening commuting patterns and the realization of a sustainable transportation in the city. Their previous research (Loo and Chow, 2008) demonstrated that providing good public transport should be connected with the new growth areas due to the increase of attractiveness of working in those areas. The improvement of public transport aims to increase connections in the new growth areas and to reduce their dependency on the city center. However, the above studies were not clearly addressing the importance of integrated space–transport development strategies in minimizing air pollution and contributing to sustainable urban transportation with a design of various urban structures. An insufficient number of studies have considered the integration of urban planning and transport systems as the reduction of air pollution. Therefore, this study focuses primarily on the design of integrated space–transport development strategies which will not only reduce congestion, but also have a significant impact on air emissions. In this study, an integrated
136
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
approach consists of a combination of space–transport development and impact assessment. The focus is on identifying a promising options based on an analysis of combinations of transport and urban land-use development. Emissions from all transport modes are traced. A comparison of choice behavior of residents and emission loads for the various alternatives provides some insight into the best strategy to increase mobility and to reduce air pollution. This study has been based on a survey to estimate the parameters in a four-step transport model, focusing on mode choice. Parameters of empirical analysis of heterogeneous traffic flow were employed in traffic simulations (Ambarwati et al., 2014a). The results of the model will be compared to residents’ preferences related to transport mode choice. In the face of rapid growth in the use of cars and motorcycles, it is necessary to consider the strategy of space–transport development in order to reduce the number of private vehicles and the high level of air pollution caused by private vehicles. In general, the motorcycle is a relatively fuel efficient but highly polluting mode. The results of this study are expected to inform the design of a suitable city structure linked to improvement of PT system in order to minimize the phenomenon of urban sprawl in developing countries and to assist in policy making with regard to addressing air pollution as part of the control of urban sprawl; a recommendation is made regarding the best way to minimize air pollution. This paper is organized as follows: data collection and methodology are elaborated in Section ‘‘Data and framework”. Background information on the public transport and urban spatial development of the city of Surabaya is also described in Section ‘‘Data and framework”. Section ‘‘The results of simulation analysis” details a simulation analysis of transport performance for a combination of urban structure scenarios and improved public transport systems. Section ‘‘Assessment of emission load for each case” discusses the estimation of emission parameters and compares the impacts of the different spatial structure and PT options. Section ‘‘Conclusions and observations” presents conclusions and recommendations for further analysis. Data and framework This section describes the data collection for the study area and the methodology which uses a combination of two models, i.e. the Java Spatial Model (JSM) and OmniTRANS. This section consists of data needed to estimate emissions, the parameters of the four-step transport model for each scenario, and the background for the set-up of the transport and spatial developments. Data The case study area is the city of Surabaya (Fig. 1), the capital city of the province of East Java, comprising 31 districts and 163 sub-districts (called ‘‘desa”) with a total area of 327 km2. The city has a population of approximately 3 million, a high density of more than 11,000 persons per km2, it can be considered a highly urbanized area (Statistic Bureau of Surabaya City, 2010).
Fig. 1. Location of study area (the city of Surabaya) in the province of East Java, Indonesia.
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
137
Build-up areas have concentrated in the current larger urban centers in most big cities in Indonesia since the 1980s, such as Jakarta, Bandung, Semarang, Yogyakarta, Medan, and Makasar, which have similar conditions and urban transport problems with the city of Surabaya. Currently, congestion and air quality have become major problems in these urban centers; this will grow out of proportion when these urban centers approximately double in size in the next twenty years. This rapid and uncontrolled growth has strong consequences for mobility for the residents of these metropolitan areas. The development of transport infrastructure is lagging behind the times, with resulting large scale congestion and air pollution. In the future, the urban development of most big cities in Indonesia should be managed a sustainable growth by controlling land use integrated with a transport network in order to decrease huge congestion and pollution, and to maximize accessibility. The spatial unit for the models is the desa, so data is collected at the desa level. The transport and spatial models require socio-economic and transport behavior data as well as associated schematizations. The 2010 Census data have been used to analyze population, employment, attractiveness variables and land-use for each of the 163 desa in the city. Transport data was collected by means of an extensive survey (2012) and the video recording of traffic. From the questionnaires distributed, a total of 554 respondents, a representative of residents of 163 desas. Analysis was based on data collected from all respondents. The questions were divided into three groups: socio-economic background, trip characteristic and transport mode choice, residents’ preferences regarding public transport and living conditions. Planning methodology The conceptual framework, presented in Fig. 2, describes the computational steps to quantify the components of integrated plans for addressing the urban sprawl phenomenon. It consists of the application of two models to simulate the performance of alternative combinations of spatial-transport development. In the present paper, the output focuses on the emissions for those combinations. The models employed in this study are OmniTRANS, to set up the transport model and to assess the environmental impact, and the Java Spatial Model (JSM), to simulate spatial settlement. The JSM was calibrated in previous research (Verhaeghe et al., 2009), while the calibration of the transport model was discussed in previous research (Ambarwati et al., 2014b). A set of simulations, comprising a spatial settlement strategy and a transport system set-up were carried out. The impacts from this set of simulation cases were compared and formed as input for the formulation of strategic options for integrated development of the city. Fig. 2 presents an overview of this methodology. A total of four spatial settlement scenarios (scenarios for the transport system) and five alternative (public) transport situations are combined into nine simulation alternatives (cases). The cases are Spatial scenarios: (1) Existing (2010) settlement, (2) Settlement 2030 based on a continuation of current trends, (3) Settlement 2030 with the consideration of a socalled compact zone in the western part of the city, and (4) Settlement 2030 with a polycentric city. Transport alternatives: (a) Existing situation (2010), (b) Situation 2030 with currently planned (current trend) transport options, (c) Situation 2030 currently planned and BRT, (d) Situation 2030 currently planned + MRT + LRT, and (e) Situation 2030 currently planned + all PT systems (BRT, MRT, LRT) Nine cases are simulated, i.e. 1a, 2b, 2c, 2d, 2e, 3b, 3e, 4b, 4e. The computations in the 3 modules can be summarized as follows: Firstly, JSM (Java Spatial Model): this model used the input of the statistics for population and employment from the 2010 census, and applied socio-economic trends and spatial planning interventions. The output of JSM (projection of population, employment, and housing area) was used to generate
Fig. 2. The conceptual framework of space–transport development.
138
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
the traffic demand which was the input for the transport model of each scenario. Secondly, OmniTRANS transport model: the input for OmniTRANS consisted of zonal data (desa), transport networks and user behavior. The zonal data was obtained from JSM and the network was set up with a GIS application. The user behavior consisted of: trip production and attraction parameters, distribution functions, BPR-functions, and VOT (value of time). The parameters of the different functions were designed in a previous study (Ambarwati et al., 2014b). In the simulation of the OmniTRANS model, residential density was inputted to determine the number of trips generated from the residential areas in each zone. The various transport modes consisted of cars, motorcycles, public transport (bus, minibus, MRT, LRT, and BRT), bicycles, and walking. Walking was considered as access and egress mode to public transport system. Thirdly, impact assessment module: computation of emissions based on the transport flows simulated with OmniTRANS; in particularly the load of emissions such as CO, CO2 , NOx, pm10, and HC. This study only focused on CO emissions. Spatial development scenarios Simulation of spatial development strategies is based on the existing spatial panning and transport network, the design of a compact zone in the particular regions related to the Spatial Masterplan of Surabaya, and design of a polycentric structure intended to shorten the travel distance. Four alternative city structures are considered with distinct spatial planning. They are elaborated below. Current situation and current trend The current situation 2010 describes the current settlement in for the city of Surabaya. The spatial schematization comprises 163 administrative zones (desas) of the city of Surabaya; 5 regions are considered (central, north, south, west, and east) as shown in Fig. 3. The total area of the city is approximately 33,637 hectares, about 65% of which can be found in the west and east regions. The built-up area in these regions is less than 15%, meaning that there is more potential development in these regions than in other regions of the city. Since the 1980s, settlement development has extended to the outskirts of the urban area, particularly on the western and eastern sides of the city. There are 135,000 households living in the suburbs. Some zones in the western regions have the highest job density due to new industrial enterprises. These regions create a significant number of jobs for the city, approximately 20% of total employment. They have a low population density resulting in a high employment-population ratio. The transport infrastructures in these regions are inadequate, and as a consequence it takes more than 90 min by motorized vehicles and more than 110 min by PT to reach the city center. The current trend 2030 comprises the projected development of Surabaya City in 2030, based on a continuation of the current situation and implementation of current planned transport interventions; further an annual growth of 5% GDP
Fig. 3. ID for the desa level and regions of the city of Surabaya.
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
139
and 1% employment has been adopted over a 20 year period (2010–2030). In this scenario, the population will reach approximately 4.5 million residents, with the highest population in the western regions. Employment growth is more equally divided over the city with a total number of almost 1.6 million jobs. The current trend 2030 includes the expansion of the transport network with the extension of eastern ring road, a new collector, and local roads in the western regions. This is expected to increase residents’ ability to access to the city center from the eastern and western regions. Compact zone strategy – balanced distribution of settlement Creating a compact zone, aiming at a high density with mixed land-uses and complementary functions such as housing, shopping and offices, is considered a sound strategy to reduce energy use in households and the extent of everyday travel in densely populated area. The central urban areas activities represented the highest level of leisure-time travel by plane (Holden and Norland, 2005). Compact zone strategy is considered in the Spatial Masterplan of Surabaya. Observation on the distribution of employment and residential density in the city indicates relatively high employment in the western part of the city attracting workers from other parts of the city. In this part of the city, the local government has set up a new factory centers, and this has led to a higher employment density in this area of the city, as seen in Fig. 4a. Fig. 4b illustrates how this part has a high employment-population ratio due to a low population density, yet maintains a high employment density. Currently, the residential areas is located the certain areas, away from industrial areas and other business areas. Under the idea of compact growth and to better manage the balance between employment and residential density, the compact zone strategy proposes the extensive development of residences, shopping facilities, and educational centers only in the western part. It means that a new sub-center in the western part would be designed to increase the number of workers living in this part and the importance of intra-zonal commuting movements. Currently, this part has less accessibility than other parts, The expectation is to reduce travel length and the number of trips to the city center from this part, as a result the effect of urban sprawl can be minimized and residences’ accessibilities can be encouraged. Such a spatial strategy is adopted in the present analysis of urban development options for the city and the consequences on transport has been quantified. The compact zone strategy in the present analysis has been designed using a similar overall current trend 2030 with GDP growth 5% and employment rate of 1%. However for this zone (the western part of the city) the ratio of employment to population increases 3% from the current trend 2030 (Ambarwati et al., 2014b). Job-housing balance (JHB) – polycentric city Reducing work trip distance and an increased use of public transport are primary goals for planners ad policy makers (Horner, 2004; Black, 2010) towards a sustainable (economic, environmental, and social) urban development. Various researchers (Ewing and Cervero, 2001; Cervero, 2002; Cervero and Duncan, 2006; Guo and Chen, 2007; Levine et al., 2012) argue that the job-housing perspective should aim to reduce the spatial disparity between homes and jobs, so raising the potentially beneficial effect on sustainability, i.e. shorter commutes and reduction in the use of private vehicles. Breheny et al. (1992) emphasized that dwellings, workplaces, public and private services should be developed around public transport nodes intended to reduce commuter distance. A design of job housing balance (JHB) is proposed in this study aimed to promote job location close to housing developments. This study plans a polycentric structure where both population and jobs are decentralized by designing 4 sub-centers in all parts of the city and associating these parts with the expansion of local PT (bus and minibus) as feeder systems. Under this urban form, intra-zonal commuting movements are expected to emerge as a new travel pattern in all parts of the city. The locations of the current main center and the 4 sub-centers are indicated in Fig. 5. The new sub-centers are designed for mixed land use, aimed at reducing residents’ dependence on the city center for their subsistence activities (working, shopping, schooling). Development of poly-centers has been carried out in order to shorten the distance between job and housing location, involving the design of new sub-centers in the different regions of the city. This strategy is also expected to reduce travel length for each region to the city center and to increase travel to the nearest sub-center. As a result, it is necessary to encourage residents to increase their use of walking, bicycling, and transit usage due to land use mix and closer travel within smaller spatial areas (new location patterns with a new sub center in each region of the city). A minimum commute concept (called MIN: minimum average commute in length) has been used to characterize a particular urban structure. It represents a jobs-housing benchmark which has its origin in urban economic theory. MIN represents a theoretical work travel pattern in which workers are assigned to alternate workplace locations, resulting in work trips with the lowest possible length. Urban sprawl disturbs the job housing balance and increases the MIN. The MIN can be quantified using a linear optimization called the transportation problem (Hitchcock, 1941; White, 1988; Giuliano and Small, 1993). The analytic framework can be formulated as follows: Minimize MIN
MINv ¼
n X m 1X cij xij W i¼1 j¼1
ð1Þ
The objective function minimizes the average travel length in the assignment of workers (xij ) between homes and jobs. Subject to
140
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
Fig. 4. Job density (a) and population density (b) in each zone of the city. m X xij ¼ Oi ;
8i ¼ 1; . . . ; n
ð2Þ
8j ¼ 1; . . . ; m
ð3Þ
j¼1 n X xij ¼ Dj ; i¼1
xij P 0 X bijv cijv cij ¼ v
ð4Þ ð5Þ
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
141
Fig. 5. Location of main center (central business districts (CBD)) and new sub-centers.
where n = the number of origin or home zones, m = the number of destination or employment zones, W = the number of workers in the study area, Oi = the total number of workers living in zone i; Dj = the total number of jobs in zone j; cij = the travel length between home site i and job site j; xij = the number of workers living in zone i and working in zone j; bij = the choice probability of mode v home site i and job site j. At the opposite spectrum of the MIN, a theoretical random commute can be considered. The assignment of workers between homes and jobs is random with an equal probability that any home will be matched to any job. To explore the ultimate potential effect of urban form for Surabaya, a MIN has been prepared associated with a polycentric city structure in which 4 sub-centers are added to the existing main center. In this development scheme, the average commute will be shortened because of the high accessibility to new jobs located in the new sub-centers and will encourage a strong expansion of local PT (minibuses) as feeder systems. The design of a polycentric urban structure with 4 sub-centers is expected to become a model which can be used by other big cities in Indonesia. Improvement of public transport (PT) systems Several needed improvements of public transport have been identified. The set up and modeling of those public transport systems, including model calibration by means of an extensive survey, have been elaborated in previous study (Ambarwati et al., 2014b). Currently, the Surabaya public transport network consists of minibusses (paratransits) and busses. There are 68 available minibus routes, and 22 bus routes (minibus accommodates 8–12 passengers, while bus has 50–55 seats). The frequency of minibusses and busses are approximately 20–25 veh/h and 5–6 veh/h respectively. The current performance of public transport such as capacity, quality, and efficiency was measured by conducting an on-board survey for bus and minibus/paratransit. The results of this survey were used as input for the design of improvements to public transport by employing the OmniTRANS model. Each part of the city is served by approximately 26 minibus routes, except the southern part of the city, which is accommodated by 10 minibus routes. The recommended improvements comprise a MRT (monorail), a LRT (tram line), and a BRT (Bus Rapid Transit) with a grid structure. For analysis, the different systems have been combined into 3 cases (Section ‘‘Planning methodology”): BRT, MRT + LRT, and all PT systems. The current local PT is associated with the improvement of the PT system as a feeder system. Fig. 6 presents the schematizations for the 3 cases. A BRT (Bus Rapid Transit) system was planned on the main roads of the city with the proposed routes (see in Fig. 6a). The BRT system should be more integrated with the road network of pedestrians to make the system more attractive to them (Fitts and Midgely, 2010). The properties of this BRT system were proposed as follows: 80 seats, 12 vehicles per hour per route, with a ticket price of about $0.51, an average speed 30 km/h, and stop spacing of 0.5–2 km. Due to the current amount of congestion occurring in Surabaya, the city of Surabaya is planning a Mass Rapid Transit (MRT) and tramline (LRT), as illustrated in Fig. 6b. The aim of the MRT is to provide a sustainable transportation system in Surabaya, to reduce the congestion and to provide alternative options of public transport that are safe, comfortable and scheduled (Department of Transport of Surabaya City, 2013). The properties of the proposed MRT and LRT are: a ticket price of approximately $0.82, stop spacing of 0.5–2 km, seat capacity of 177 seats for the MRT and 100 seats for the LRT, frequency of 12 veh/h, and an average speed of 55 km/h for the MRT and 35 km/h for the LRT.
142
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
Fig. 6. BRT-system with grid structure (a), development of MRT and tram lines (b), and all PT-systems consisting of BRT, LRT, MRT and feeder systems in the future for Surabaya City (c).
143
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
Based on integration of the previous ideas, improvement of all PT systems consists of MRT (monorail), LRT (tram line), BRT (Bus Rapid Transit) and buses and minibuses as feeder systems (see in Fig. 6c). The improvements to public transport are expected to encourage the users of cars and motorcycles to shift their transport mode to the public transport systems. In Section ‘‘Planning methodology” nine system cases have been defined comprising a combination of a spatial structure (at a particular time: present or 2030) and a public transport option. Each of these cases has been simulated. The simulated flow patterns for the different modes of transport are used to compare the significance of the improvement of various PT systems linked to changes of land use patterns. The comparison involves assessing the performance of the different cases as well as the emission loads associated with these transport patterns. This comparison is made in order to reveal the extent of the change of spatial or transport system development intended to increase residents’ mobility and to reduce air pollution. Section ‘‘The results of simulation analysis” discusses the transport performance. Sections ‘‘Assessment of emission load for each case” and ‘‘Conclusions and observations” analyse the emission loads.
The results of simulation analysis Important features of the transport system, concerning accessibility and air quality are modal split and commuting distance. Alternatives which promote a modal shift towards public transport and decrease commuting distance will positively contribute to increased accessibility and air quality. Below, Section ‘‘Modal split” discusses the effects of the nine simulation cases on modal split, and Section ‘‘Commuting distance” presents the results on commuting distance.
Modal split Improvements in public transportation systems are intended to increase the people’s use of public transport in each case. Each case here examines the change of modal split for each vehicle class. Table 1 presents the results of modal split for these different cases. The following points can be noted from the chart: – An improvement of PT has only a slight effect on the mode share of PT (C2 versus C5); it reduces slightly the use of private cars and the share of motorcycles. – The compact zone spatial intervention has on its own little effect on the mode split for PT (C2 versus C6). The alternative design of a compact zone with an added improvement of PT encourages this mode split (C2 versus C7); it can be observed that C7 has practically no effect on the motorcycle mode, which remains by far the dominant mode. – A drastic change in urban structure (job housing balance (JHB)), C9 has a drastic effect on the mode split for PT: the PT mode share increases to 70%. The alternative is associated with a strong increase in the bicycle mode. The mode shares of motorcycles and private cars reduce drastically. In particular the strong reduction in motorcycle use is remarkable because it remains dominant in all other cases. An extensive reduction of use of the motorcycles in some zones, such as zones in the south, west and east sides of the city occurs because of accommodated by MRT and BRT lines. This can be explained by the strong improvement of home-job trip situation and the attractiveness of the new centers: the demand for trips in this urban structure matches very well with residents’ preferences of use of public transport in Surabaya. – Case C9 also corresponds to the preference of residents for a short trip distance, as communicated through the survey. Forty percent of respondents respond that they prefer to use PT for trips of 8.5 km or less. This means that change modal split with enlightened spatial development, i.e. polycentric city planning with designing 4 new sub-centers in the city increase in users of PT. In short, Case C9 demonstrates the importance of urban structure on accessibility and obviously represents part of an ideal solution for development of the urban structure of Surabaya. Provision of sufficient PT is necessary to take over from other modes but an adequate urban structure is needed to make a new PT system successful.
Table 1 Modal split for each scenario (%). Transport mode
PT Motorcycle Car Bicycle 1
Spatial development scenarios
Transport options
Cases (C1) current base
(C2) current trend
(C6) compact zone
(C8) JHB1
(C3) BRT
(C4) MRT and LRT
(C5) all PT
(C7) all PT + compact zone
(C9) JHB + all PT
5.82 58.73 34.88 0.57
5.62 63.39 30.23 0.75
5.66 64.3 29.2 0.82
62.48 23.3 13.02 1.23
5.69 60 28.86 5.48
7.72 58.7 28.7 5
8.51 59.8 26.58 5.04
9.23 60.3 25.39 5.09
70.66 10.35 6.43 12.57
Includes a strong expansion of localized PT (small buses).
144
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
The design of the city structure is aimed to reduce travel to the city center and to minimize residents’ dependence on the city center, as explained in the next section.
Commuting distance In this study, we assessed the minimum average commute (MIN). The values were required to evaluate the significance of the space–transport development strategies in shortening the trips. Table 2 presents the average commuting distance for the different cases. The following observations can be made: – In Cases C1 and C2, a high number of commutes take place to the city center, particularly by motorcycle. – The case with improvement of public transport (C5) has had a significant effect on reducing the average commuting distance of cars in contrast to motorcycles, meaning that what used to be long car trips could instead be trips making use of PT. – The compact zone case C7 has a considerable effect on shortening the commuting distance. – The polycentric urban structure and improved PT systems, seen in case C9, results in the lowest commuting distance for all transport modes; the commute distances are about half of the present (C1) and future current trend (C2) cases. It is clear that case C9, with a very high PT mode share and short travel distance, will have a strong effect on residents’ mobility. This means that C9 is expected to have an effect on the emission load. The emission load is elaborated next in S ection ‘‘Assessment of emission load for each case”.
Assessment of emission load for each case Unit emission loads (g/veh km) are vary for different modes and different speeds. Relevant emission substances are CO, CO2 , NOx, pm10, and HC. An analysis of the air quality of the city of Surabaya indicates that CO emission is dominating the air pollution problem because of the extensive use of motorized vehicles. Most motorcycles emit CO, NOx and HC, while cars exhaust 30–50% CO and HC (Badami, 2005). Chiou and Chen (2010) also revealed that HC and CO emission levels are primarily determined by car and motorcycle exhaust. Therefore the analysis of this research will focus on CO emissions. The weighted emission values for each vehicle type depend on the average standard speed of each vehicle class for each road type (motorway, urban, and rural). The values were analyzed using the procedure for determining air emission rate for each vehicle class at the micro level as stipulated in the guideline of The Ministry of Public Work (1999). The emission rate ((g/veh km) is expressed as
qCO ¼ 867:92 U 0:8648 v
ð6Þ
where U v is average speed standard for each vehicle class for each road type in km/h. The emissions for a particular traffic flow over a particular stretch of road by a particular vehicle class are determined by
Air pollutant ðgÞ ¼ traffic flow of each vehicle class ðvehÞ emission value for each parameter for each type of vehicles ðg=veh kmÞ road length km
ð7Þ
The CO emission load has been computed by simulation of the OmniTRANS model, based on the simulated flows on the transport network, for various cases; the results are presented in Table 3. The following can be noted:
Table 2 Average commute for each alternative (in km). Average commute
PT Motorcycle Car 2
Spatial development scenarios
Transport options
Cases (C1) current base
(C2) current trend
(C6) compact zone
(C8) JHB2
(C3) BRT
(C4) MRT and LRT
(C5) all PT
(C7) all PT + compact zone
(C9) JHB + all PT
4.08 5.52 4.95
4.8 5.78 5.52
3.44 3.94 3.57
2.28 3.05 2.6
3.83 6.17 4.34
4.5 5.33 3.49
3.65 5.43 3.46
2.89 3.44 2.89
2 2.57 2.22
Includes a strong expansion of localized PT (small buses).
145
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146 Table 3 The average CO emission for all cases (in tons/peak-hour). Transport mode
2010
2030
Spatial scenarios
Car Motorcycle PT Total
PT system options
(C1)
(C2)
(C6)
(C8)
(C5)
(C7)
(C9)
22.8 53.4 84.3 160.5
65.2 124 45.1 234.4
70.2 118.9 22.7 211.8
1.4 4.8 10.9 17.2
19.8 43.9 19 82.7
14.2 29.3 8.5 52.8
0.4 0.8 9.6 10.9
– For the present (C1), the estimated peak hour load of CO from cars, motorcycles and PT for the whole city is around 160 tons. This is expected to increase dramatically in the future (C2) to 234 tons if no spatial and PT interventions are implemented. This will be primarily caused by an increase in the use of private vehicles. These findings are confirmed by Cervero (2000), who explained that exposure levels and thus health risks are lower, on the other hand tailpipe emissions and fossil-fuel consumption will continue to increase. – Improvement of the PT system (C5) has little effect on decreasing the extents of load of private vehicles (car and motorcycle), as reflected in a slight reduction of CO emissions. – The design of a compact zone (C6) does not provide a significant reduction in emission load. – The hypothetical polycentric city structure and expansion of PT systems (C9) results in a drastic change in the emission load (90% reduction), due to the combined effect of short commuting distances and the very strong transition to public transport. In comparative analysis of all alternatives, C9 yielded, as expected, the lowest level of air pollutant concentrations (CO emission), average commute, and the highest number of passengers using public transport system. Because of high housing development in the suburbs, the design of new sub-center in each region should be undertaken and connected to the city center by public transport systems. The real obstacles to achieving that sort of development pattern are cost, coordination between the authorities for housing and job developments. Realization of the development pattern should be supported by regulation of spatial development and management of land use. The regulation should be recommended to the local government of Surabaya.
Conclusions and observations The results of the analysis indicate that an integrated approach, comprised of interventions in urban structure and public transport, is necessary to address the mobility and emission problems in fast growing urban areas and the effects of urban sprawl. It does not seem that an improvement of public transport would substantially reduce the use of motorcycles, in which residents currently see as the best way to deal with congestion. It therefore at present becomes the dominant transport mode (59.8%) and is projected to remain so until 2030. It is estimated that the compact zone approach in the current City Masterplan will have a modest impact on PT mode share: from 6% to 9%. This situation will result in a significantly shortened commute distance down to 38%, and a substantial reduction in emission load, down to 60%. Testing a polycentric urban structure for the city of Surabaya with 4 sub-centers, an associated minimized commuting distance, and an improvement of public transport, has suggested that the mode share of public transport could be increased from 6% to 70%, the commuting distance could be shortened by 53% and the emission load could be reduced by 90%. The polycentric structure shortens trips and makes public transport much more attractive to users, resulting in a travel pattern matching users’ preferences. The test with the polycentric city concept indicates that considerable improvements in transport performance and air quality are potentially possible by changing the whole land use in the city. The proposed polycentric structure and associated minimization of commute distance has the potential to decrease the motorcycle mode share to 10%. Improvement of public transport is generally considered to be the main tool for improving accessibility and reducing the emission load. On the basis of the present analysis, it can be concluded that public transport will only be really successful if associated with a suitable urban structure. This means that the key to minimizing urban sprawl is an improvement of the public transport system linked to a change in the city’s urban structure. Integration of spatial-transport strategies as well as an environmental assessment would provide long-term advantages by designing improvements of public transport systems related to settlement development. If carried forward, the integrated approach should also be based on residents’ preferences regarding the way in which these systems are developed and implemented. Further study should assess the feasibility of changing the direction of growth in the city, particularly with a consideration of residential location along transit corridors, and the influence of urban structure on transit capacity due to the change of modal split for PT.
146
L. Ambarwati et al. / Transportation Research Part D 44 (2016) 134–146
References Ambarwati, L., Pel, A.J., Verhaeghe, R., van Arem, B., 2014a. Empirical analysis of heterogeneous traffic flow and calibration of porous flow model. Transp. Res. Part C: Emerg. Technol. 48, 418–436. Ambarwati, L., Verhaeghe, R., Pel, A., van Arem, B., 2014b. Investigating the effects of improving public transport system linkage to spatial strategy on controlling urban sprawl: evidence from Surabaya city, Indonesia. WIT Trans. Built Environ. Badami, M.G., 2005. Transport and urban air pollution in India. Environ. Manage. 36, 195–204. Banister, D., 1996. Energy, quality of life and the environment: the role of transport. Transp. Rev. 16, 23–35. Black, W.R., 2010. Sustainable Transport: Problems and Solutions. Guilford. Breheny, M.J., Association, R.S., et al, 1992. Sustainable Development and Urban Form, vol. 2. Pion, London. Cervero, R., 2000. Transport and Land Use: Key Issues in Metropolitan Planning and Smart Growth. University of California Transportation Center. Cervero, R., 2002. Built environments and mode choice: toward a normative framework. Transp. Res. Part D: Transp. Environ. 7, 265–284. Cervero, R., Duncan, M., 2006. Which reduces vehicle travel more: jobs-housing balance or retail-housing mixing? J. Am. Plan. Assoc. 72, 475–490. Chin, A.T., 1996. Containing air pollution and traffic congestion: transport policy and the environment in Singapore. Atmos. Environ. 30, 787–801. Chiou, Y.C., Chen, T.C., 2010. Direct and indirect factors affecting emissions of cars and motorcycles in Taiwan. Transportmetrica 6, 215–233. Chiou, Y.C., Wen, C.H., Tsai, S.H., Wang, W.Y., 2009. Integrated modeling of car/motorcycle ownership, type and usage for estimating energy consumption and emissions. Transp. Res. Part A: Policy Pract. 43, 665–684. Costabile, F., Allegrini, I., 2008. A new approach to link transport emissions and air quality: an intelligent transport system based on the control of traffic air pollution. Environ. Model. Softw. 23, 258–267. Department of Transport of Surabaya City, 2013. Surabaya MRT: Surabaya Mass Rapid Transportation (SMART). Technical Report. Surabaya, Indonesia. Duncan, J., 1989. The Search for Efficient Urban Growth Patterns: A Study of the Fiscal Impacts of Development in Florida. Publisher not identified. Ewing, R., Cervero, R., 2001. Travel and the built environment: a synthesis. Transp. Res. Rec.: J. Transp. Res. Board 1780, 87–114. Fitts, A., Midgely, P., 2010. An Overview of Urban Mobility and the Implementation of an Area Traffic Control System in Surabaya. Frumkin, H., 2002. Urban Sprawl and Public Health. Public Health Reports 117, 201. Giuliano, G., Small, K.A., 1993. Is the journey to work explained by urban structure? Urban Stud. 30, 1485–1500. Guo, J.Y., Chen, C., 2007. The built environment and travel behavior: making the connection. Transportation 34, 529–533. Hitchcock, F.L., 1941. The distribution of a product from several sources to numerous localities. J. Math. Phys 20, 224–230. Holden, E., Norland, I.T., 2005. Three challenges for the compact city as a sustainable urban form: household consumption of energy and transport in eight residential areas in the greater Oslo region. Urban Stud. 42, 2145–2166. Horner, M.W., 2004. Spatial dimensions of urban commuting: a review of major issues and their implications for future geographic research. Prof. Geogr. 56, 160–173. Levine, J., Grengs, J., Shen, Q., Shen, Q., 2012. Does accessibility require density or speed? A comparison of fast versus close in getting where you want to go in us metropolitan regions. J. Am. Plan. Assoc. 78, 157–172. Loo, B., Chow, S., 2008. A framework to study urban spatial structure and sustainable transport: minimum commuting, excess commuting and the potentials of greater transport sustainability. In: A seminar Paper Presented at the Hong Kong Geography Day 2008, Hong Kong. Loo, B.P., Chow, A.S., 2011. Jobs-housing balance in an era of population decentralization: an analytical framework and a case study. J. Transp. Geogr. 19, 552–562. Marquez, L.O., Smith, N.C., 1999. A framework for linking urban form and air quality. Environ. Model. Softw. 14, 541–548. Martins, H., 2012. Urban compaction or dispersion? An air quality modelling study. Atmos. Environ. 54, 60–72. Mitchell, G., 2005. Forecasting environmental equity: air quality responses to road user charging in leeds, uk. J. Environ. Manage. 77, 212–226. Nejadkoorki, F., Nicholson, K., Lake, I., Davies, T., 2008. An approach for modelling CO2 emissions from road traffic in urban areas. Sci. Total Environ. 406, 269–278. Statistic Bureau of Surabaya City, 2010. Surabaya City in Figures. Technical Report. Surabaya, Indonesia. Stone, B., 2008. Urban sprawl and air quality in large US cities. J. Environ. Manage. 86, 688–698. The Ministry of Public Work, 1999. Procedure for Prediction of Air Pollution in Micro scale from Traffic. Technical Report. Jakarta, Indonesia. Law no.017/T/ BM/1999. Verhaeghe, R.J., Zondag, B., Grashoff, P.S., 2009. Space and Water: Development of the Java Spatial Model. Technical Report. Minister of Economic Affairs. White, M.J., 1988. Urban commuting journeys are not wasteful. J. Polit. Econ., 1097–1110