Modeling and visualizing the spatial uncertainty of moving transport hubs in urban spaces - a case study in NYC with taxi and boro taxi trip data
Urban human mobility relies on the presence of transport hubs, which connect different modes of transport, public and private. This paper proposes a technique for relating taxi trip origins and destination hotspots for gaining knowledge on the spatial uncertainties of transport hubs, more precisely their movements within specific times. The outcomes of applying the techniques are matter of further investigation of spatial uncertainty perception, representation, and visualization. In the stages of the approach, the outcomes of transport hub movements are related with more general functional transport regions resulting from NYC public transport services.
Urban environments are characterized by complex and dynamically changing human mobility. Vehicle movement trajectories of urban vehicle fleets, as taxi fleets, can help extracting this knowledge. Taxi trip data is usable in many different applications of different research domains.
Due to often-immense sizes of the data extracts, many applications make only use of taxi trip origin and destination points. In an aggregated view, these points may be clustered into origin and destination hotspots, which are connectable, especially, when there are different taxi fleets with specific operational restrictions. This promotes the automated deduction of local knowledge on smaller scales of investigation.
Understanding time-dependencies and dynamics of moving transport hubs is beneficial for planning, decision-making, and various ongoing research in connection with smart cities and intelligent transportation systems (ITS).
The methods of the proposed technique include density-based point clustering, convex hull estimations, polygon matching, and intersection. The components are part of an extendible workflow. The aim of the mentioned polygonization of point clusters is, besides providing visibility at certain map scales, representing positional accuracies together with uncertainties in density-based connectivity of trip origins and destinations.
Matching between different types of areal polygons relies on the Dimensionally Extended nine-Intersection Model (DE-9IM) as introduced by Clementini et al. (1994).
The outcomes of applying the techniques are matter of further inspections and visual analysis. The visualization of the polygon matching results aims to show the spatial uncertainty of boro taxi trip destination hotspots for at least four successive time windows.
Besides the resulting boro taxi trip destination hotspots, there are also extractable yellow taxi trip origin hotspots and functional transport regions resulting from public transport stations of subway and bus lines. This has a semantic connection to the restricted yellow zone and specific commuting patterns between Manhattan and the outer boroughs of NYC. The further steps include matching these spatiotemporal hotspots with characteristic transport hubs, and testing a specific visualization scheme for representing spatial and spatiotemporal uncertainties of the mentioned transport hubs.
Clementini, E.; Sharma, J. & Egenhofer, M. J., 1994: Modelling topological spatial relations: Strategies for query processing. Computers & Graphics, 18 (6), 815-822.