Exploring the changes of individuals' travel behavior in response to COVID-19 and their influencing factors based on mobile phone data
Shuli Zhou,Suhong Zhou,Fengrui Jing,Luhui Qi,Jianjun Li
DOI: https://doi.org/10.1016/j.jth.2024.101788
IF: 3.613
2024-03-23
Journal of Transport & Health
Abstract:Introduction The COVID-19 pandemic has produced a profound impact on travel behavior. Yet, few studies explored how individuals resist such impact to maintain their daily life from the perspective of people's responses. A comprehensive theoretical framework to elucidate individual travel behavior changes and a quantitative analysis of the pivotal factors driving these changes was lacking. Methods This study constructed a framework for individuals' travel behavior in response to COVID-19 and utilized anonymized mobile phone trajectory data before and during COVID-19 in Guangzhou to explore the changes of individuals' travel behavior on workdays. Gradient boosting decision trees was used to quantify the degree of the factors driving travel changes. Results During the pandemic workdays, non-essential trips experienced a reduction of 51.4%, while essential trips decreased by 30.94%; Subway trips decreased by 1.63%, whereas motorized trips increased by 8.75%. The most significant decrease was observed in travel duration (50.92%), followed by travel frequency (41.28%), and travel distance (39.19%). The three travel indices displayed significant social disparities and spatial disparities. However, the relative relationships among these three travel indices did not vary with groups and spaces. Risk perception and intervention measures, emerged as primary drivers of individuals' travel changes, contributing to 27%–35% of the total impact, while individual socio-economic characteristics and built environment accounting for less than 10%. Implications The results indicated that individuals curtailed both their essential and non-essential trips and adjusted their travel mode in response to the epidemic. These changes in travel purpose and mode led to differing reduction rates in travel frequency, duration, and distance. These findings contribute to individual travel changes prediction and fine-grained epidemic modelling during other public health emergencies like COVID-19.
public, environmental & occupational health,transportation