Scholarly article on topic 'Activities Relocation for a Sustainable Mobility System'

Activities Relocation for a Sustainable Mobility System Academic research paper on "Civil engineering"

CC BY-NC-ND
0
0
Share paper
Academic journal
Transportation Research Procedia
OECD Field of science
Keywords
{"activity location" / "land use model" / "location problem" / "transit system;"}

Abstract of research paper on Civil engineering, author of scientific article — Adrienne Brandi, Stefano Gori, Marialisa Nigro, Marco Petrelli

Abstract The paper analyzes an existing Land Use Transport Integration (LUTI) model in terms of its capacity to generate sustainable mobility behaviours in an urban area promoting the use of the public transport. The cited LUTI assumes the residual capacity of the mass transit system as a key element to indicate location and intensity of new residential and activities developments. It appears as a tool for the adoption of integrated transport-land use policies, especially for the short term planning. A systematic and accurate analysis of this LUTI is reported considering a real case of study (the city of Rome, Italy): five scenarios with different characteristics both in terms of supply network and travel demand have been considered. Finally, suggestions are reported to overcome some weaknesses that can arise in practical applications of the LUTI.

Academic research paper on topic "Activities Relocation for a Sustainable Mobility System"

Available online at www.sciencedirect.com

ScienceDirect

Transportation Research Procedía 5 (2015) 4-12

Transportation

Procedía

www.elsevier.com/locate/procedia

SIDT Scientific Seminar 2013

Activities relocation for a sustainable mobility system

Adrienne Brandi, Stefano Gori, Marialisa Nigro*, Marco Petrelli

Department of Engineering, Roma Tre University, 62 Via Vito Volterra, Rome 00146, Italy

Abstract

The paper analyzes an existing Land Use Transport Integration (LUTI) model in terms of its capacity to generate sustainable mobility behaviours in an urban area promoting the use of the public transport. The cited LUTI assumes the residual capacity of the mass transit system as a key element to indicate location and intensity of new residential and activities developments. It appears as a tool for the adoption of integrated transport-land use policies, especially for the short term planning. A systematic and accurate analysis of this LUTI is reported considering a real case of study (the city of Rome, Italy): five scenarios with different characteristics both in terms of supply network and travel demand have been considered. Finally, suggestions are reported to overcome some weaknesses that can arise in practical applications of the LUTI.

© 2015TheAuthors.Publishedby ElsevierB.V. Thisis an open access article under the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/4.0/).

Peer-review under responsibility of the Societa Italiana dei Docenti di Trasporti (SIDT). Keywords: activity location, land use model, location problem, transit system;

1. Introduction

The structure of the city clearly influences mobility behaviors and the activity location is one of the main factor determining population trips. On the other hand, the transport system plays a very important role in accessing to these activities: transport supply affects the activities location choices, moving the economy of the city, its settlement structure and, consequently, the social environment. Then, it is clear that land use and transport system development are strictly connected and there is an increasing need to integrate them in order to achieve a more sustainable environment.

* Corresponding author. Tel.: +39-06-57333632; fax: +39-06-57333441. E-mail address: marialisa.nigro@uniroma3.it

2352-1465 © 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/4.0/).

Peer-review under responsibility of the Società Italiana dei Docenti di Trasporti (SIDT). doi:10.1016/j.trpro.2015.01.012

Land use and transportation have been heavily dealt with by the research community, especially at urban scale. Different approaches are adopted, some of them focusing on single land use models or on single transportation models: in the first case the future transportation system is assumed to be fixed, while in the second case is the land use to have a fixed spatial pattern (Oryani and Harris, 1996). More interesting are definitely the approaches based on the analysis of the interaction between the land use and the transport system and on the role of the land use characteristics in the travel patterns choice.

Models trying to integrate transport planning and land use have been formulated since 1960. According to Timmermans (2003), they are usually divided into three generations:

1. first generation or aggregate spatial interaction-based models: they are based on aggregate data and on the principles of gravitation and entropy maximization (Lowry, 1963, Garin, 1966, Mackett, 1983);

2. second generation: multinomial logit models based on the principle of utility-maximization (Echenique, 1994, de La Barra, 1989);

3. third generation: based on micro-data and activities travel patterns (Miller and Savini, 1998, Veldhuisen, et al, 2000).

The major limitations of previous studies are in the limited capacities of representing the behaviour of the decision-maker and in the simulation models: instead, the problem is very complex and thus difficult to model. Moreover, the most advanced models require a tremendous amount of data to be calibrated and validated, while the investment, especially for that models designed to give short-term forecasts, are quite poor. Thus, it seems that the utility of integrated land use - transport models is mainly in supporting long term, strategic planning decisions (Timmermans, 2003).

Following the transit oriented cities (Newman and Kenworthy, 1996) developed around the main transit stops, the idea of maximizing access to public transport generated the concept of Transit Oriented Developments (TODs). TODs represent a significant research area in order to promote the adoption of sustainable transport mode. An extensive debate concerns also the role played by land use variables as the population and activities densities to explain the level of car and public transport use. Sinha (2003) demonstrates, with the collection of different data from 46 cities in United States, Australia, Canada, East Europe and Asia, that a high urban population density seems to be a primary element to increase transit ridership. Eidlin (2005) highlights from the analysis of the cities of Los Angeles, New York and San Francisco that the critical issue to link the use of public transport to such variables is not in the average density value, but in its distribution within an urban area. Mees (2009), from the comparison between urban densities and transport mode shares of Australian, Canadian and United States urban areas, highlights that variations in density are little or no correlated to transport modes share, which seems more closely related to different transport policies. Gori et al. (2012) reports about the importance of increasing density of residences and activities, but also as single actions on these variables are not successful due to the complexity of the problem integrating the land use to the transport system

Recently, Gori et al. (2013) proposed a Land Use Transport Integration (LUTI) model, which refers to the concept of TOD: specifically, it adopts the residual capacity of the mass transit system (existing or its short term configuration) as a key element to define the location and the intensity of new residential and activities developments in urban areas. However, compared to similar models whose applications are usually analyzed in a qualitative manner, pointing out a series of experiences and proposing different policies, it is set as a practical tool to promote the use of public transport.

This study starts from this last scientific contribution and from the investigation of the solution method proposed. Different scenarios for the real case study of the city of Rome (Italy) have been considered in order to apply the LUTI considering different hypotheses in terms of supply network and travel demand. The aim is to stress the capability of the procedure to promote sustainable travel behaviours. Finally, suggestions are reported to overcome some weaknesses that can arise in practical applications of the LUTI.

2. Evaluation of an existing LUTI model

The approach proposed by Gori et al. in 2013 has been here considered as an interesting reference point, since the idea is to overcome the temporal discrepancy between infrastructural actions on land use and infrastructural actions on the mass transit systems, trying to restore the correct connection between land use and the public transport system. To reach this objective, the authors stated that it is fundamental to match the expected land use of the different urban zones with the potential offered by the existing transport system or by its short-term development. The residual capacity of the mass transit system (existing or its short-term development) is hence considered as a key element that can indicate the location and the magnitude of new residential and activities developments, so that the resulting trips are just directed to use this remaining capacity, promoting sustainable mobility behaviors.

The method starts from the estimation of the potential transit demand (in terms of generated and attracted trips) induced by the growth of new activities developments close to the links of the mass transit system with available residual capacity. Then, this potential transit demand is converted into location and intensity of residential and business volumes.

The estimation of the potential transit demand is formulated as an Origin-Destination Matrix Estimation problem (ODME, Cascetta 2009, Cipriani et al., 2011), consistent with the capacities of the links of the mass transit system network and formalized as:

minE?=1 w "(V « - si f=1^=1 Pij xt yt Tjj )2 + y¿T*.))2 (1)

where:

Va = maximum desired flow on link a

Tjj* = demand value from origin i to destination j

Pij = fraction of the demand from i to j using link a

wa, z = external weights in order to give more importance to the demand term or the flow term s = scale factor to give more importance to a specific flow value xi, yi = maximum level of variation for origin i and destination j N = number of zones

The first term in (1) indicates the difference between the expected link value and its actual value, while the second term adds the difference between the starting demand matrix and the final one. Therefore, it is looking for a demand matrix not too far from its starting configuration and leading to link values as closest as possible to the expected values.

In the analyzed approach, the maximum desired flow on link a, Va, is the capacity of the mass transit system, while the output of the procedure is the transit demand matrix xyTj. Once obtained xyTj, the changes in trip generation and attraction trips can be evaluated respectively by the xi and yj variables. Obtaining xi and yj variables is essential to pass to the second phase of the method: in fact they are the main inputs in order to trace back the demand models (trip generation and attraction models) and to determine possible changes in the development of residences and activities for each existing zone of the study area.

The optimization problem reported in (1) is the most generic formulation to be adopted to compute the potential demand; starting from (1), different hypotheses can be introduced:

1. if the second term in (1) is deleted, the procedure has high possibility to strongly modify the starting demand value, giving rise to high variation of the xi and yj variables;

2. it is possible to consider, as maximum desired flows on links, a reduced capacity computed considering the real accessibility to the different stops of the mass transit system.

The evaluation of the solution method is carried out in Gori et al. 2013 adopting a qualitative investigation of the results: by the way, it was not possible to investigate if such results are comparable with sustainable mobility behaviors, i.e. if the new generated trips are consistent with mobility behaviours oriented to the use of the public

transport system. One of the objectives of the present study is to investigate this issue. 2.1. Case study and Scenarios definition

To verify the effectiveness of the solution method proposed by Gori et al. 2013 to promote sustainable mobility behaviours, it has been applied to the existing mass transit network of the city of Rome, focusing only on the metro network.

The urban area of Rome is actually characterized by a population of 3 million with 1.1 million employees, contributing to about 552,000 trips in the morning peak hour. There are two metro lines (A, B with its branch called B1) extending for a total of 45 km. These lines have a radial structure with the main interchange in the city centre (Termini rail station). Actually, about 11 km of the metro network (24%, Fig.1,a) present a value of the passenger flow higher than the capacity (maximum saturation degree equal to 1.48), while the remaining extension of the network presents residual capacity, especially in the peripheral links (Fig.1,b).

Fig 1. Overflows (a) and residual capacity (b) of the metro network in Rome, Italy

Several scenarios have been proposed to better understand the mechanism of the method especially during the ODME phase (Fig.2):

1. Scenario 1: the residual capacity of each link of both the two metro lines has to be reproduced by the procedure;

2. Scenario 2: the residual capacity of each link of only one metro line (line A) has to be reproduced by the procedure;

3. Scenario 3: the residual capacity of only two peripheral links of only one metro line (line A, between Baldo degli Ubaldi stop and Valle Aurelia stop, and between Subaugusta stop and Giulio Agricola stop) has to be reproduced by the procedure.

These three scenarios have been run without considering any constraints on the starting transit demand both in terms of its total value and its distribution. Then, results are expected to modify completely the structure of the current demand matrix.

In order to bind more this potentially strong impact on the potential demand, other two scenarios have been formulated where constraints on the total generated trips by each zone are added: these constraints have been inserted as new terms inside the objective function of the ODME next to the values of the transit capacities to be reproduced.

Fig 2. Metro lines/Segments considered in the different scenarios

The expected result is in such a case a variation of the demand expressed only in terms of a redistribution of travel between destinations:

1. Scenario 4: the residual capacity of each link of both the two metro lines and the generated trips of each zone have to be reproduced by the procedure;

2. Scenario 5: the residual capacity of only two peripheral links of only one metro line (as in Scenario 3) and the generated trips of each zone have to be reproduced by the procedure.

2.2. Scenarios evaluation

The different scenarios have been simulated using the EMME software (Florian, 2014), where the ODME phase has been run using the Spiess procedure (Spiess, 1990): it permits to perform the transit demand estimation according to the known passenger link flows. In this application, the passenger link flows are set equal to the capacity of the considered transit links, while for Scenario 4 and 5 also the transit generated trips of each zone are considered as fictitious "monitored passenger link flows".

Respect to the starting transit demand value of about 240,000 passengers/hour, all the scenarios bring to an increase of demand between 10% and 20% (Table 1).

Scenario 1 is one of the most constraint scenarios, since the ODME procedure has to reply the capacity of all the links of the two metro lines. In order to reach this goal, the procedure tends to increase the transit demand, especially all the OD components inside the basin of the two metro lines: this is the reason why the change in transit demand is much higher inside the Municipality, where the metro network develops, respect to outside (+3%). Analyzing the details of this increment of transit demand for each single urban zone of the city, it is possible to underline as the areas closest to the terminals of the metro lines have the highest increments of demand, while the central areas can have also a decrease of demand: it happens because the most external sections of the metro lines have high values of residual capacity, while the central sections are over capacitated.

Scenario 2 reports similar results in comparison to Scenario 1: the highest difference is in the total change that is less evident respect to Scenario 1 (+9.4% respect to +16.2%, Table I), since the need to reply the capacity is required for only one line (line A).

Scenario 3 is much less constrained respect to the previous two scenarios because in such a case it is required to reproduce only the capacity of two links (4 transit segments): the increase in the total transit demand value (+15.5%)

is mostly due to the change in demand outside the Municipality (+17.4%), since the two links are peripheral respect to the lines development. Moreover there are no reductions of demand in some urban zones, but only increments are generated: it happens because there are no more capacity values to be replied in the central area where the network is already over capacity, so the ODME procedure is free to increase the demand to reply only the capacities of the two peripheral links. It does not care about further increase of the overflows.

The higher freedom of the ODME procedure in case of Scenario 3 is also demonstrated by the convergence of the objective function of the ODME (Fig.3), that happens after only 5 iterations respect to the 35-40 iterations required in the other scenarios.

Scenarios 4 and 5 are introduced to bind more the total demand value, but the results report however an increase of the total transit demand with values similar to those obtained respectively for Scenario 1 and 2: it happens because the values of capacities to be replied are numerically much higher than the values of generated trips by zones, so the ODME procedure trusts mainly on their values.

Table 1. Change in Transit demand and travelled distance after the ODME phase using residual capacity

Analyzed Scenario

Change in total transit demand [%]

Change in transit demand inside the Municipality [%]

Change in transit demand outside the Municipality [%]

Change in average travelled distance [%]

Scenario 1 Scenario 2 Scenario 3 Scenario 4 Scenario 5

+23.1 +12.2 +14.5 +32.2 +14.7

+15 +10 +10 +21 +16

Fig 3. Objective function trend of the ODME phase for the first 3 scenarios.

About the distribution of trips generated by the ODME procedure, it is possible to underline as Scenario 1 and 4 lay to generate a strong increase of trips transversal with respect to the development of the two metro lines, since in such a way it is possible to cover with the generated flows both the lines. Instead, in the other scenarios the trips longitudinal to the metro lines, whose links are considered in the ODME procedure, are most common.

Finally about the average travelled distance, respect to the actual value of 10 km, the application of the ODME procedure brings to an increase of 10 to 20%: it means that on an average trip of 10 km, the increment of distances due to the new location of residences and activities can reach also 2 km.

In summary, although the idea behind the analyzed approach seems to be promising, the adoption of the ODME procedure to find the new values of generated and attracted trips by zones, given the residual capacity of the mass transit system, generates some weaknesses:

1. an increase of the total transit demand and the inability to bind its starting value also using the total generated trips by zones;

2. an increase of transversal movements that, for the current configuration of the metro lines adopted as a case study, need to high travel times consisting of a waste of time in the central interchange (Termini station);

3. in some cases, an extreme increase of the overcapacity;

4. the increase of average travelled distances.

Especially points 2 and 4 are not in accordance with sustainable issues and then with a new configuration of the land use able to promote travels with public transport.

3. Revising the LUTI model

In order to overcome the previous underlined weaknesses, starting points for the improvement of the LUTI are here reported.

Maintaining the main concept of searching for a specific value of residual capacity reduction, the method could involve only the relocation of activities (no new developments) without making use of the ODME procedure.

Moving only existing activities avoids variations in total travel demand entity, while not using the ODME permits to generate only transit-oriented behaviour.

>| AC a: 01 |-_

AC End

Fig 4. Main steps of the proposed model for activities relocation.

Specifically the LUTI could work following the flow-chart reported in Fig. 4 and specifically:

1. Point 1: urban zones with high attraction, but with a low transit accessibility are identified;

2. Point 2: urban zones close to the links of the mass transit system, with residual capacity and availability to host the activity volumes, are selected;

3. Point 3: a share of activity volumes are relocated from the urban zones of point 1 to the urban zones of point 2;

4. Point 4: the relocation of activity volumes has to interact with the transport demand in order to take into account current and resulting (after the relocation) transit accessibility, as well as the distribution of trips respect to the configuration of the mass transit network (to be sure of the use of the public transport system after the relocation);

5. Point 5: check of constraints on both the land use and the transport system side;

6. Point 6: evaluation of the residual capacity of the mass transit system after the relocation: if it decreases more than a threshold set at the beginning of the procedure, the procedure stops, otherwise a new iteration is performed restarting from point 1.

Input of the model are in such a way: current activities intensity and location; accessibility measures to the final destination points; residual capacity of the mass transit system; available space and/or urban constraints for urban areas where activities have to be relocated; private and public transport travel demand; the public transport network. All these input are usually at the disposal of Urban Planning Administrations, which makes the LUTI easily applicable.

Applications of this "modified" LUTI are needed in future developments of the study to validate its capability to overcome the founded shortcomings and to be a useful support system to suggest activities relocation pursuing the goal of sustainability for a short term horizon.

4. Conclusions

This study starts from the analysis of an existing LUTI model and from the investigation of the solution method adopted: the interesting idea behind the analyzed LUTI is in the adoption of the residual capacity of the mass transit system (existing or its short term configuration) as a variable to indicate location and intensity of new activity developments in urban areas. The solution method adopts an Origin-Destination Matrix Estimation (ODME) where the usually adopted traffic counts have been substituted by the capacities of the links of the mass transit system. Then, in this paper firstly different scenarios have been proposed in order to apply the LUTI with the aim of stressing the capability of the procedure to promote sustainable travel behaviours: results showed different weaknesses of the procedure, mainly related to the ODME and to the lack of mobility and urban constraints.

Then, the study proposed a way to improve the model in order to overcome the previous underlined weaknesses and specifically: 1) it involves only the activities relocation, no new developments are considered, then avoiding distortion in travel demand entity; 2) it does not make use of the ODME, then avoiding no transit-oriented behaviours.

References

Cascetta, E. (2009) Transportation systems analysis: models and applications. Springer, New York (NY), USA

Cipriani, E., Florian, M., Mahut, M., Nigro, M. (2011), "A gradient approximation approach for adjusting temporal origin-destination matrices".

Transportation research. Part c, emerging technologies, vol. 19, p. 270-282, 2011, ISSN: 0968-090X, doi: 10.1016/j.trc.2010.05.013 De la Barra, T. (1989), Integrated Land Use and Transport Modelling, Cambridge, Cambridge University Press

Echenique, M.H. (1994), Urban and regional studies at the Martin centre: its origins, its present, its future, Environment and Planning B, 21, 517533

Eidlin, E. (2005) The worst of all worlds, Transportation Research Record: Journal of the Transportation Research Board, No.1902, pp.1-9 Florian, D. (2014). Emme 4 Overview: New features, demonstrations and upgrade details, INRO, Montreal, Quebec, Canada 2013-11-14, INRO

Model City Seminar in Kuala Lumpur Garin, R.A. (1966), A matrix formulation of the Lowry model for intra-metropolitan activity location, Journal of the American Institute of Planners, 32, 361-364

Gori, S., Nigro, M., Petrelli, M. (2012) The impact of land use characteristics for sustainable mobility: the case study of Rome, European Transport Research Review (2012) 4:153-166

Gori S, Nigro M, Petrelli M (2013). A new method to recover the correct land use and public transport interaction. C.A. Brebbia, Urban Transport XIX - Urban Transport and the Environment in the 21st Century. WIT TRANSACTIONS ON THE BUILT ENVIRONMENT, vol. 130, p. 279-290, Southampton:WIT Press, ISBN: 978-1-84564-716-2, ISSN: 1743-3509, Kos, May 2013, doi: 10.2495/UT130221

Lowry, I.S. (1963), Location parameters in the Pittsburgh model, Papers and Proceedings of the Regional Science Association, 11, 145-165

Mackett, R.L. (1983), The Leeds Integrated Transport Model (LILT), Supplementary Report 805, Transport and Road Research Laboratory, Crowthorne

Mees, P. (2009) Density and transport mode choice in Australian, Canadian and US cities, 32nd Australasian Transport Research Forum Auckland, ATRF

Miller, E.J. and Savini, P. (1998) The Integrated Land Use, Transportation, Environment (ILUTE) Modelling System: A Framework, Proceedings 77th Annual Meeting of the Transportation Research Board, Washington, D.C.

Newman PWG, Kenworthy JR (1996) The land use-transport connection. Land Use Pol 13(l):l-22.

Oryani, K. and Harris, B., 1996. Enhancement of DVRPC's Travel Simulation Models Task 12: Review of Land Use Models and Recommended Model for DVRPC. Report Prepared for Delaware Valley Regional Planning Commission, URS Consultants, Inc., Philadelphia

Sinha, K.C. (2003) Sustainability and Urban Public Transportation, Journal of Transportation Engineering, pp.331-341

Spiess, H. (1990). A Gradient Approach for The O-D Matrix Adjustment Problem , EMME/2 Support Center, Aegerten, Switzerland 1990-06-15, 5th International EMME/2 Users' Conference.

Timmermans, H. (2003). The Saga of Integrated Land Use-Transport Modeling: How Many More Dreams Before We Wake Up?. 10th International Conference on Travel Behaviour Research Lucerne, 10-15. August 2003

Veldhuisen, K., Timmermans, H.J.P. and Kapoen, L.L. (2000), Ramblas: A regional planning model based on the micro-simulation of daily activity travel patterns, Environment and Planning A, 32, 427-443