This research investigates how pedestrians choose routes around a subway station during the evening peak hour in an attempt to discover which factors affect pedestrian decision-making. An ArcGIS based model was built to predict pedestrian walk patterns and was adjusted based on a field survey of actual pedestrians. The adjusted prediction model indicates that pedestrians on the streets around Central Station are mainly those employees in Cambridge who work within a 1km walking distance to the station, but live remotely. Pedestrian route choice is affected by the walking distance, and pedestrians are more likely to walk directly to the station without too many detours, especially when the weather is terrible. Combing urban network analysis in the ArcGIS platform and a field survey calibration approach, this research reveals the movement of people at the neighborhood scale, which will enable urban planners to more efficiently implement people-oriented designs and planning policies.
Key words
Pedestrian Walking Simulation, Gravity Based Model, Urban Network Analysis