Spatiotemporal structure of taxi services in Shanghai: Using exploratory spatial data analysis

Admitting the stochastic nature of taxi movement at the individual level, this paper explored the spatiotemporal structure of taxi services in Shanghai from a macro perspective. A GPS-based taxi service dataset was first processed to derive a total of 9921 taxies' OD, which were then tallied by individual street districts. An index known as Traffic Ratio Density was computed to characterize the level of taxi services for each street district and facilitate the mapping of its spatiotemporal variation. In the end, the method of Exploratory Spatial Data Analysis was used to identify spatial clusters of taxi services over time. Both global and local Mo ran's Ivalues were computed for Shanghai as a whole and for individual street districts. The positive values of the global index strongly suggested high and stable concentration pattern across all the time-periods. The local index showed that the taxi OD pattern has a high-high cluster in the CBD area, versus the low-low cluster in the suburban regions, and between them was a stochastic distribution. There was no noticeable temporal variation at eitherglobal or local level, indicating arather stable spatial and temporal structure of taxi service distribution.