This paper examines network analysis for urban areas. The research is focused on
the problem of deÞnition and visualisation of network geography and network spaces
at different scales. The urban scale of analysis is examined and different spatial indices
are considered. The built environment of the city is considered as a reference
environment for a road network density index. The latter is implemented in order
to study the spatial interactions between network phenomena and spaces and to
provide further elements for the analysis of urban shape. The study is focused in
particular on understanding spatial patterns drawn by networks and in helping with
the delimitation of city centres. Different approaches are used to obtain the two
indices: a grid-based analysis and a spatial density estimator based on Kernel Density
Estimation. The two methodologies are analysed and compared using point data for
the urban road network junctions and street numbers as house location identiÞers
in the Trieste (Italy) Municipality area. The density analysis is also used on road
network junctions data for the city of Swindon (UK) in order to test the methodology
on a different urban area.
