Consideration of the historical developoment of towns shows that their scale has been set by the prevailing travel speed, but the present ubiquity of the private automobile means that urban development is perhaps less sensitive to individual change in the transport network than it has been in the past. The basic trade-off between the desire for personal space and good accessibility between to a range of desire activities (a range which tends to grow with time as affluence, transport and technological progress expand the possibilities) is understood qualitatively, but because prediction involves the summing of opposing tendencies, we need a much better quatitative understanding than we have at present. This paper will review briefly the type of modeling work which attempts to provide this understanding by representing the two-way interaction between land-use patterns and the transport system, mentioning work from a number of different countries. Most of the models in use are based on the Lowry formulation, though with added mechanisms at varying levels of sophication and aggregation. They have provided this understanding considerable food for thought, but their predictive abilities are largely untested and probably unreliable. New approaches are being developed, from the `self-organising' urban models whit their appealing ability to break spatial symmetry to the `budget' models of travel behavour which have important implications for the way activity patterns (and in the long term, therefore, urban form) are likely to be affected by transport changes. But the very long response-time of urban development renders the proving of all these models problematic, a point which is underlined by the empirical finding that planning policies impose the most important constrains on how land-use will be affected by transport. The observations that most people are willing to trade potential reductions in travel time for a wider choice of travel destinations illustrates the dangers inherent in any simple-minded approach to `performance indicators' for urban systems. The intricate trading between a variety of desirable but conflicting attributes that, if the `quality' of a town can be measured at all, a whole host of different indicators will be required. Moreover, if the variety of choice is important in itself, the trading rate between one attribute and another will not be a fixed quantity. (A)
Samenvatting