Understanding the spatiotemporal response of dockless bike-sharing travel behavior to the small outbreaks of COVID-19

Abstract

While the impact of the COVID-19 pandemic on human mobility is widely studied, the effects of small outbreaks remain underexplored. This study addresses this gap by investigating spatiotemporal behaviors of dockless bike-sharing (DBS) over small outbreaks. First, descriptive statistics and network theory are applied to identify the spatiotemporal evolution across different outbreak phases. The Prophet model is then introduced to evaluate short-term responses, and the copula theory to analyze behavioral preferences. Results indicate increased count and decreased duration during small outbreaks, contrasting with broader pandemic observations. Although DBS activity is most localized during the outbreaks, it expands eastwards into wider urban peripheries, showing community merger and growth. Notably, network communities reveal distinct spatiotemporal behaviors compared to grid-based analysis, offering new perspectives for targeted strategies. Travel behaviors vary with rising cases within a week: trip count initially increases, then declines in the urban core, while trip distance and duration shorten in over half of the region. Probabilistic modeling, however, suggests that greater distances are more likely on days following new cases in western Shenzhen, highlighting the need for comprehensive analysis. This research provides multi-level spatiotemporal insights into bike-sharing behavior for stakeholders in urban and transport planning.