Repository logo

Incorporation of micro-level analysis in strategic urban transport modelling: with a case study of the Greater Beijing



Change log


Deng, Bin 


Many developing countries and regions are suffering from severe urban transport problems arising from accidents, congestion, air pollution, rising carbon intensity, and chronic under-funding of infrastructure and services. The problems make those cities the most polluted and often the least liveable. Strategic transport modelling has been recognised as an effective approach for developing and testing policy options, especially where it is integrated with land use planning and urban design. However, in most developing-country cities strategic transport modelling has been out of reach for practical policy use because of its sophisticated data and skill requirements, which currently imply unaffordable high costs and long durations for model development. This means that strategic urban transport modelling is the least available where it is needed most urgently. Meanwhile, the spread of smart data in mapping and urban activity monitoring has often been just as rapid in developing countries as in the developed. This has triggered new approaches in micro-level analyses of transport networks, personal movements and vehicles. In the most advanced cases, the new analyses have started to influence strategic modelling. The main hypothesis of this dissertation is that an incorporation of the micro-level smart data and analyses in strategic urban transport modelling will make it feasible to establish a sufficiently robust strategic transport model for evidence-based policy analysis with cost, time and skill thresholds that are close to being affordable in developing country cities. In order to test this main hypothesis, a number of novel model development tasks have been carried out which contribute to the field of applied urban modelling. This new approach aims to contribute to the transformation of the prevailing modus operandi where model development could not start in earnest until extensive data collection and skills training have been completed to a situation where a sufficiently robust model can be established cheaply and quickly to support on-going and incremental refinements. More specifically, new modelling tools have been developed as part of this dissertation using sparse GPS taxi traces to identify slow-moving and stopping traffic hotspots using an extended density-based spatial clustering algorithm that is tolerant of significant data noise, and to estimate congested road speeds (which used to be very costly and time-consuming to obtain if at all). The micro-level network, congested speeds and insights into the nature of the congested traffic have been incorporated into a MEPLAN-based strategic transport model interacting with a MEPLAN-based land use and travel demand model. This means that the strategic economic, social and environmental impacts of transport interventions can be tested in a robust way through accounting for the interactions among transport, land-use and background social-technical trends. A new approach to establish the medium to long term visions for alternative travel demand management and transport investment scenarios has been tested using this model. The methods and algorithms have been tested in a case study of the Greater Beijing region, which consists of the municipalities of Beijing and Tianjin together with the surrounding areas in the province of Hebei. The government’s data regulations of restricting overseas studies to using only publicly available data sources have made the case study ideal for testing the new approach. The potential of the new strategic urban transport model has been tested through a wide range of policy scenarios. The results suggest that the new approach developed in this dissertation has made it not only cheaper and faster to develop a robust model, but could also potentially fill a gap in the lack of medium to long term perspectives regarding major road and metro investments over the next two decades. Such analyses could be of critical importance in improving the performance of the transport system in terms of safety, economic efficiency, air quality and carbon reduction given the long lead times to plan and deliver transport infrastructure investments.





Jin, Ying


Strategic transport modelling, Taxi GPS data, Transport policy intervention, Scenario tests, Greater Beijing Region, Data mining, Urban modelling


Doctor of Philosophy (PhD)

Awarding Institution

University of Cambridge