The robustness of human-driven urban systems
 

    The city is a human-driven self-organizing structure. Most modern models of urban development ignore this obvious property, and substitute humans with the derivatives of their activity: infrastructure elements. In infrastructure-based models human population is considered as the fast component of the system, which adapt to slower infrastructure changes.
    Despite noteworthy achievements, the infrastructure-based view of urban systems is limited conceptually. A more “proper urban model”, for example one that simulates urban sprawl, is necessarily a multi-agent one, where decisions to resettle or leave on the one hand or to develop, destroy, renovate or change the land use on the other, are made by interacting agents of different (but not many) types: householders, development and commercial companies, local councils, municipalities. In this way, the achievements of cognitive, behavioral and social sciences are directly incorporated into the urban model. Furthermore, multi-agent models can serve as verification tools for infrastructure-based models, and can be used to establish the limits of their application.
    To elaborate the human-based approach, we present a model of the urban dynamics that includes agents of two types, householders and dwelling developers, and investigate the evolution of a simulated urban system, based upon different assumptions of agents’ behavior. The model clearly demonstrates that spatial gradients of dwelling prices, self-organizing at the initial stages of urban evolution, are very robust; their changes demand strong and extended intervention at above-agent levels. General implications of this result for the urban dynamics theory are considered. Model results are compared with data on Tel Aviv land-use dynamics over the last fifty years.
    A spectrum of urban multi-agent models, all including agents of two or more different kinds, is presented, and urban phenomena that could be investigated with these models are discussed.


Itzhak Benenson
University Tel Aviv
Department of Geography and Human Environment
http://eslab.tau.ac.il
bennya@post.tau.ac.il

Erez Hatna
University Tel Aviv
Environment Simulation Laboratory, Porter School of Environmental Studies
http://eslab.tau.ac.il
erezh51@post.tau.ac.il