The urban sprawl dynamics: does a neural network understand the spatial logic better than a cellular automata?

Figure 8 - The surface growth for each land use

and, surprisingly, along the roads; this result is unexpected because, as said earlier, the
record does not include information about road accessibility.

Moreover, the new residence seems to be attracted by the proximity to other activities
(industry and commerce). Indeed, this spatial feature characterizes the urban quality of
the Italian historical cities and villages and holds particularly in this territory.

New productive settlements will be based mainly around the existing large industrial
areas, showing a location criterion mainly driven by agglomeration economies. Such
behaviour characterizes also the larger new settlements predicted from the NN for
commerce, which, in the considered area, are clustering around new development poles
near the highway and far from the urbanised areas.

On the contrary, small and diffused expansions in industry and commerce will take
place in and around the existing urban areas, improving the urban mix and urban
agglomeration economies.

What is worthwhile to note is that the SANN findings, in terms of spatial patterns, are

Figure 9 - the predicted residential growth at 2008 (grey tones)

15

<<   <   12   13   14   15   16   17   18   >   >>