Spatially constrained virtual-physical coupling in info-epidemic coevolution: an agent-based modelling framework

As information technology increasingly intertwines physical and virtual spaces in modern cities, understanding info-epidemic coevolution is crucial for effective epidemic control. While microscopic Markov-chain models on multilayer networks are widely used, many info-epidemic frameworks treat virtual and physical networks as topologically independent and do not explicitly represent individuals’ locations, activity places, or community-level spatial heterogeneity. To address this gap, we introduced a spatially explicit agent-based model that represents individuals’ geospatial attributes and activity places within the coevolutionary process. We further proposed a spatially constrained multiplex random graph generation algorithm to model the coupling between physical contact and virtual connection networks. Parameterised by real-world survey data from a megacity, this algorithm constructed a spatially explicit info-epidemic transmission network. The model integrated rules governing individual behaviour and cross-space interactions. Simulations showed that virtual-physical coupling could enhance epidemic suppression even when it did not produce the fastest information diffusion. They also showed that information diffusion displayed spatial clustering that intensified under epidemic feedback. This framework clarifies how virtual-physical interdependence shapes spatiotemporal diffusion and supports spatially explicit analysis of info-epidemic coevolution.