Geoscience Reference
In-Depth Information
Chapter 20
Modelling and Simulation
George L.W. Perry
Introduction
Models and the practice of modelling have been the subject of ongoing debate in
geography. Modelling 'has arguably become the most widespread and infl uential
research practice in the discipline of geography, as indeed within the sciences more
generally' (Demeritt and Wainwright, 2005, p. 206). The geographical literature is
replete with reviews of various approaches to modelling and debates as to the merits,
or otherwise, of modelling itself (e.g., see Macmillan, 1989b; Canham et al., 2003;
Wainwright and Mulligan, 2004). In this chapter, I aim to provide a picture of the
'state-of-the-art' in the modelling of human-environment interactions, with a focus
on simulation models and their evaluation. The focus is on the place of models and
the nature of modelling as intellectual activities, rather than on the mechanics of
model-building. The chapter is divided into two broad sections; the fi rst focuses on
current perspectives on modelling in geography and the second uses a series of case
studies to illustrate how modelling is being practised.
Fundamental concerns for effective model-building and analysis are: (i) ensuring
that the entity under investigation is appropriately represented and (ii) obtaining
the data required to parameterise the models. These two issues relate to some of
the crucial decisions of model-making: how detailed should a model be? How much
causal (process) representation does it need to incorporate? At what scales in time
and space should it operate? The problem of determining optimal model complexity,
in terms of representational and empirical adequacy, is a recurrent theme of this
chapter. A second underlying theme is that of complexity and complexity science
(Medd, 2001; O'Sullivan, 2004). Over the last decade interest in and insights from
'complexity' and 'complexity science' have led to signifi cant shifts in modelling
socio-environmental systems. It is important to distinguish between complicated and
complex systems. In complicated systems many components interact in a linear, or
somehow predictable, manner (e.g., a multi-component, yet inherently predictable,
system such as an aeroplane), whereas complex systems may comprise but few
components, but (indirect) interactions between those components result in unex-
