ABSTRACT
Urban systems are dynamic systems and hence their analysis requires predictive models that explicitly incorporate the element of time. Consequently, the treatment of time as to whether it is discrete or continuous is fundamental in the development of predictive models. The urban stock is a major component of the spatial system. In fact the spatial structure of the city has been said to be the outcome of two interdependent elements: the locations of stock, and activities. The stock-generation model generates the stocks of houses which also form inputs for the residential model. The urban stock consists of adapted spaces such as land and buildings, and 'Channel spaces', such as the transportation network. The approach obviates the theory problem of aggregation as well as retaining the modest data requirements for empirical development. The predictive model consists of six major submodels, two of which generate urban stock, another two allocate this stock, while the remaining two are activity allocation sub-models.
