[starlogo-users] city modeling with starlogo and cellular automata

[Kent Paul Dolan]

All,

I thought you'd get a kick out of this excerpt from
an announcement of a presentation local to Arizona
State University, 2006/02/09. Starlogo proves useful
yet again.

xanthian.

Prof. Daniel Czamanski, City and Regional Planning,
Technion Institute of Technology, Israel, Director
of the City Complexity Research Lab, and Chairman of
Czamanski Ben Shahar and Co.

The evolution of urban spatial morphology is
discontinuous in space and non-uniform in time. As a
result, precise descriptions are elusive. In a
series of previous papers we depicted, analyzed, and
endowed with at least partial explanations, the
dynamics of the Tel Aviv morphology based
exclusively on historic 2D data of urban footprints
and on socio-economic time series and cross-sections
data. In this paper we begin to depict and analyze
the spatial and temporal processes of the Tel Aviv
3D urban morphology.

To this end we present a rudimentary 3D cellular
automaton simulation model in which the city's shape
is referenced by means of temporal changes in the
volume of buildings, height distribution of
buildings and their spatial distribution. The model
uses a 2-D grid. The objects in the cells are the
buildings that have variable height. Each of these
cells has a neighborhood of eight cells, called
"Moore's neighborhood".  The model describes a city
growing from flat with zero average height to one
with agglomerations of skyscrapers. In the heart of
the model there are four parameters: Initial
building share, Inertia, interaction with neighbors
and noise. Those parameters are very similar to the
matching parameters in Batty [1998].

The simulations are carried out using Starlogo
software, and experiments are conducted with
systematic variation of the four parameters. The
results indicate that cities can experience
different path growths.  Cities can be in
convergence path or in divergence course and this
depends on the setting of the parameters in the
model. Very interesting evidence is the existence of
a sharp borderline between the two zones of path
settings. The model also supports changing the
values of parameters in time; those changes in
conditions reveal that in the synthetic environment
of this lab, cities can change their behavior
drastically and even experience phase transitions.