Modeling monthly flows of global air travel passengers: An open-access data resource

Abstract

The global flow of air travel passengers varies over time and space, but analyses of these dynamics and their integration into applications in the fields of economics, epidemiology and migration, for example, have been constrained by a lack of data, given that air passenger flow data are often difficult and expensive to obtain. Here, these dynamics are modeled at a monthly scale to provide an open-access spatio-temporally resolved data source for research purposes (www.vbd-air.com/data). By refining an annual-scale model of Huang et al. (2013), we developed a set of Poisson regression models to predict monthly passenger volumes between directly connected airports during 2010. The models not only performed well in the United States with an overall accuracy of 93%, but also showed a reasonable confidence in estimating air passenger volumes in other regions of the world. Using the model outcomes, this research studied the spatio-temporal dynamics in the world airline network (WAN) that previous analyses were unable to capture. Findings on the monthly variation of WAN offer new knowledge for dynamic planning and strategy design to address global issues, such as disease pandemics and climate change.