Diffusion processes in multilayer transportation networks: the flight of the Coronavirus

Abstract

At the end of December of 2019, a new type of coronavirus, now called COVID-19 started spreading in Wuhan, China and later throughout the world. Due to the global emergency state the World Health Organization declared and the need to know more about the danger Mexico is in, we worked on analyzing the risk of the COVID-19 importation to Mexico through the Air Transportation Network with a multilayer network approach. Based on the data obtained from the public data bases of OpenFlights, we created a multiplex network in which nodes represented airports, flights represented links, and airlines represented layers. We then simulated the propagation of the coronavirus using an unbiased random walk model with probability p=1 of infection once the random walker steps in a certain airport. We found the COVID-19 spread behavior the first month is anomalous (subdiffusion) and later behaves as a normal diffusion. We also found the risk of importing the virus to Mexico increases linearly over time and after approximately one year, there is almost a 90% probability of being infected. However, it is important to mention this high risk is due to contagions by people from other countries (not China) which have already confirmed cases of coronavirus. We concluded the risk of importing the COVID-19 to Mexico is almost ineludible over time unless effective medical interventions are imposed.