Land Use and Transport Interaction Models - Where is the limit?

Land Use and Transport Interaction (LUTI) models are precious tools to integrating the many impacts and feedbacks of the location of activities on transport infrastructures and vice versa. In that sense, applying LUTI models is key to delivering regulation and planning options for urban and transport sustainability. For they have been used in practice to guide urban planners and help transport policy since the 1960’s and the pioneering work of Lowry (1964). The effectiveness of LUTI models as decision support tools is generally well accepted by transport and planning researchers and by practitioners, despite sailing through troubled waters between consultancy secrets, politically led options, data problems, rule of thumb calibration, and model openness and transparency.

At the turn of the millennium, LUTI models have developed from aggregate zone-based models to micro representation of space and disaggregated representation of agents, following increased computing capacity and availability of better GIS and individual data. As demonstrated by Wegener (2011) this modeling shift however goes with costs that impede empirical validation and further adoption in planning. LUTI modelers have long recognized that the different processes within LUTI models act and interface at different speeds (Wegener, 1986). The further granularity accompanying disaggregation then leads to dynamics that are trickier to handle. As argued by Anas (2013), LUTI models need clearer definitions and a stricter use of urban economics concepts. Likewise, we argue in this paper that LUTI models also need to take better care of geographical knowledge and spatial biases. They require the analysis of the robustness of model outcomes to the choice of spatial units and MAUP, which is analyzed by Jones et al. (2013) but also the effect of changing urban system boundaries, which is under focus here.

Defining the limits of a coherent study area for modeling is actually questioning the delimitation of a city or urban region. This is obviously not a new question to geographers and economists but it impacts deeply on how inner stocks (population, firms, …) and external flows (traffic, labor,…) are modeled in LUTI models and therefore on their outcome. From intuition and practice, we hypothesize that there is strong inertia in LUTI models outcomes because the many parameters and variables that are present in these models are eventually strongly constrained by the geographical structure considered (monocentric, polycentric, including exurbs or not, etc.). Where the outer limit of a model is traced not only impacts the internal components of the city but also questions the problem of the limits between two cities, which is not trivial.

First, we perform a meta-analysis of recent LUTI applications in European contexts based on 19 peer-reviewed articles. The lack of definition of the study area is striking. Interestingly in the first LUTI implementation, Lowry (1964) explicitly mentioned the use of an estimate of the commutershed of Pittsburgh for the next 20 years. The lack of explicit choice in later literature with models of increased complexity, stresses the need for guidelines that could improve practice in order to improve the comparability of applications and the generalization of results where possible. Second, we perform simulations on a synthetic city system using UrbanSim (Waddell et al., 2003). We gradually vary the spatial limits of the system from an inner center monocentric system to a polycentric city region. We also vary population and employment endowments, hence commuting patterns. Our simulations show that LUTI results are highly impacted by the change of limits and therefore suggest a reason why LUTI models are sensitive to large parametric shocks only. Our paper confirms that the absence of a strict theoretical rationale for city delineation weakens the effectiveness of LUTI models.

References

Anas, A. 2013. A response to the guest editorial: economics as the science for urban modeling. Environment and Planning B, 40 (6), 955 – 958

Jones, J., Peeters, D and Thomas, I. 2013. On the Influence of Scale on Urban Planning Evaluations by LUTI models. ASRDLF Congress 2013.

Lowry, I.S. 1964. A model of metropolis. Memorandum RM 4035 Rand Corporation, Santa-Monica. 136p

Waddell, P., Borning, A., Noth, M., Freier, N., Becke, M. and Ulfarsson, G. 2003. Microsimulation of Urban Development and Location Choices: Design and Implementation of UrbanSim. Networks and Spatial Economics, 3 (1), 43-67

Wegener, M., Gnad, F., Vannahme, M. 1986. The time scale of urban change. In Hutchinson, B and Batty, M. (Eds), Advances in Urban Systems Modelling. North-Holland, Amsterdam, 145–197.

Wegener, M. 2011. From macro to micro - how much micro is too much? Transport Reviews, 31, 161–177