Chinese Aviation Network Analysis under Actual Node Attack

During the Corona Virus Disease 2019 (COVID-19) pandemic, many airports were semi-closed to prevent the virus from spreading, which greatly impacted the Chinese aviation network. The current approach to the analysis of the Chinese aviation network completely removes nodes to study node attacks on the Chinese aviation network. However, it is not practical to close the airport completely while ignoring other airlines' adjustments. This paper investigates the changes of Chinese aviation networks under the real node attack and provides useful information and practical suggestions on how to deal with the transmission of infectious diseases by airlines in the future. First, we collect the flight schedules for representative dates from VariFlight, delete useless information, split layover flights and count the number of flights in each airport. Next, we compute the properties of the Chinese aviation network under node attacks such as betweenness centrality and PageRank. Finally, the properties are compared with those of normal dates to investigate the changes of the Chinese aviation network under node attack. The results indicate that the Chinese aviation network is a small-world and scale-free network. Our analyses demonstrate that the decline of airlines in hub airports can influence the properties of network and adjacent airports. Meanwhile, the node attack on a hub airport with many substitutes failed to achieve the expected result. Therefore, traffic restriction should also be applied to the substitutes if there are many large cities around the infectious city.