Unveiling urban traffic accessibility patterns and phase diagrams of traffic direction through real-time navigation data in Beijing

Urban transportation accessibility plays a crucial role in assessing traffic conditions and gaining insights into urban development. Current research on accessibility patterns often relies on sensor data, focusing predominantly on specific locations or times. Recognizing the need for a more holistic study that considers the interconnected impact of both geographical location and time variables, this research conducts a spatial-temporal analysis of over 846,720 real-time locationto -location navigation data from Baidu Maps in Beijing. The findings reveal four distinct traffic accessibility patterns: the Weekend pattern (W -pattern), Southern weekday pattern (S -pattern), Northern weekday pattern (N -pattern), and Holiday pattern (H -pattern). These patterns exhibit spatial-temporal distribution characteristics and scale invariance. Significantly, scale invariance emerges as a key feature, suggesting potential phase transitions in the dynamic change process of the traffic system. To capture this phenomenon, a new indicator is introduced, utilizing the relative velocity sign to resemble the spin direction in the Ising model. Phase transition -like occurrences are identified through the phase diagram of the traffic state. These observations may provide useful insights into the geographic and temporal patterns of transportation accessibility in growing metropolitan areas.