The Impact of Human Mobility on Spatial Disease Dynamics

Understanding human mobility is crucial for modeling the spatial spread of human infectious diseases. The quantitative description of spatial epidemics is based on two prominent theoretical approaches, diffusive dispersal and direct coupling or effective force of infection. The first ansatz assumes random walk movement of the host between different locations where as the second employs an effective force of infection between distinct populations. Both models are inconsistent with important aspects of human mobility, most importantly the bidirectional movements between individuals´ homes and distant location. We introduce and investigate a novel epidemiological model that explicitely takes into account this bidirectional nature of human movements. In various topologies (networks and lattices) we find significant differences as well as similarities among all three models, depending on the parameters. On a lattice we obtain an analytical expression for the velocity of the propagating epidemic front. In contrast to the diffusion approach, our model predicts a saturation of the velocity with increasing traveling rate. Our analysis is supported by numerical simulations on both lattices and networks and provides a framework for incorporating the abundance of pervasive data on individual human mobility into disease dynamics modeling.