Evaluation of automated street network modelling shows reliable analytical outcomes in urban contexts, suggesting efficient urban analysis methods.
Key Points
This research aims to compare automated street network modelling against a hybrid approach for centrality and accessibility analysis in various urban settings.
Evaluated street network models in four cities: Nicosia, London, Gothenburg, and Madrid.
Analyzed geometric segmentation, node densities, and street segment lengths.
Examined correlations of angular integration and betweenness between models.
Performed spatial autocorrelation analysis on varying thresholds.
Automated models showed higher node densities and shorter segments compared to hybrid models.
Strong rank correlation in angular integration between both modelling approaches, especially at larger spatial radii.
Moderate to high correspondence in angular betweenness, sensitive to local changes.
Divergence in accessibility metrics increased at larger thresholds due to finer connectivity in automated networks.