Urban Modelling: Course Outlines

Lectures to be Given by Professor Michael Batty

January 2017: Chinese University of Hong Kong

Lecture 1: Urban Modelling and Simulation

This first lecture will set the context for analytical and predictive modelling of city systems. The notion that we can predict the future is examined and discussed within the wider context of our previous attempts at doing so by building mathematical/symbolic models of cities which are implemented using various types of computation. Different types and styles of models are defined, from descriptive to predictive to prescriptive and these are then illustrated using various kinds of social physics. This serves to introduce notions about time and distance and in this, we will focus on the concept of gravitation which is key to many kinds of urban simulation. We will have conclude with some issues concerning the development of models to inform rather than to predict, to explore ‘what-if’ types of scenario and to provide a framework for planning support systems.

Lecture 2: Gravitation and Spatial Interaction

In this second lecture, we will start with gravitation and potential and derive a family of such models, introducing the framework of entropy maximising which has analogues in discrete choice theory and utility maximizing. We will discuss the four stage transport model and then introduce ways of extending such models to deal with land use, housing markets, and economic activities such as employment and retailing. This provides the basis for land use transportation interaction (so-called LUTI) models and we will develop basic structures for such models and recount the way these have developed over recent years, being applicable to large scale metropolitan forecasting and planning

Lecture 3: Integrated Urban Models

We will continue from our discussions in Lecture 2 and introduce basic model structures in particular the simplest due to Lowry, UrbanSim, MEPLAN, DELTA,  MARS, the IRPUD model, ILUTE and so on. We will also nest these models in their wider context of demographic and economic forecasting and introduce various methods of integrated assessment that involves stitching different models together. We will conclude with a discussion about how these models might be used in decision-support and how they relate to different time scales of forecasting from diurnal to very long term decanal cycles.

Lecture 4: Predicting the Impact of Large Scale Urban Infrastructures

We will define different kinds of infrastructure and discuss the role of infrastructure planning within the wider urban and regional planning process. Infrastructure has become ever more significant in western countries due to the need for extensive regeneration of the existing physical stock of building and transportation as well as the way in which national and regional governments are using infrastructure to revive their flagging economies which have barely recovered since the great recession hit a decade ago. In the UK, new transport infrastructure is a major item on the planning agenda and in this talk, I will show how we can use variants of our urban land use and transportation models to explore the impact of infrastructure on land and development as well as on the kinds of transport we use to move around our cities and country. The model we have developed for the entire country (England and Wales) is called QUANT. It is an online forecasting model focused on three modes of transport – bus, rail and road. We will illustrate its use in evaluating very large infrastructure projects such as the new high speed rail lines planned for the UK, linking London to the regions, fast rail in London, and the development of new airport capacity in SE England. The focus will be on how travelers shift their modes of travel under such scenarios, thereby making their trips more efficient in terms of travel time and how new activities locate in an around such major travel improvements.