Title: Epidemic risk perception and social interactions lead to awareness cascades on multiplex networks

Abstract: The course of an epidemic is not only shaped by infection transmission over face-to-face contacts, but also by preventive behaviour caused by risk perception and social interactions. This study explores the dynamics of coupled awareness and biological infection spread within a two-layer multiplex network framework. One layer embodies face-to-face contacts, with a biological infection transmission following a simple contagion model, the SIR process. Awareness, modelled by the linear threshold model, a complex contagion, spreads over a social layer and induces behaviour that lowers the chance of a biological infection occurring. It may be provoked by the presence of either aware or infectious neighbours. We introduce a novel model combining these influences through a convex combination, creating a continuum between pure social contagion and local risk perception. Simulation of the model shows distinct effects arising from the awareness sources. Also, for convex combinations where both input sources are of importance, awareness cascades that are not attributable to only one of these sources, emerge. Under these conditions, the combination of a small-world face-to-face and a scale-free social layer, but not vice versa, make that the extent of the infections decreases with increasing transmission probability.