From Axial to Pedestrian Path-Centre Line: The Case of 3D Pedestrian Network in Hong Kong, Central

Abstract

In urban areas where road network and pedestrian network are intimately linked yet wildly different, axial lines and road centre line geometries are often used as strategic pedestrian networks to evaluate accessibility, pedestrian flow potential against empirical pedestrian volume counts. However, pedestrian networks in multi-level urban environments are usually separated from the road network. In Hong Kong, the pedestrian network combines both steep topographies as for the road network and separation from the road network due to extensive publicly accessible multi-level outdoor and indoor pedestrian network. Taking multi-level urban environment of Central area in Hong Kong, probably one of the most complex urban area in the world as a case study, we examine the correlation coefficient between spatial configuration and pedestrian volume. The study mainly focuses on comparison between three kinds of network data generalisation models: axial line map, segment map derived from axial map, and pedestrian path centre-line map, the link/node standard. Results show that pedestrian network segment and path centre-line representations can capture built environment features better than axial generalisation. A 3D indoor-outdoor centre-line map is also constructed to represent the intricate multi-level pedestrian network. The results suggest that only a floor by floor version of the Indoor-outdoor model would provide an adequate measure of potential pedestrian flow through multi-level urban environments. It is also showed that Angular properties of central area have the strongest effect on pedestrian route choice in the study area. A calibration of 50% - 50% Angular - Euclidean hybrid metric gives a stable result in terms of interpreting pedestrian movement potential.