Epidemic Dynamics on a Geospatial-Temporal Network Using a Mobility-informed SEIR Model

We use a mobility-informed SEIR model with an exportation-driven force of infection to model disease spread using a geospatial-temporal municipality-level mobility network derived from telecom origin–destination data. The model captures both intra- and inter-municipality transmission and incorporates scenarios with external seeding and parameter variation. Simulations conducted using the mobility network of Surigao del Norte, Philippines reproduce overall epidemic trends and reflect the probable magnitude of low case detection. Compared to an equivalent scale-free random network, the simulations highlight the central role of mobility hubs in driving earlier and more intense outbreaks. Model validation using reported case data shows that simulations driven by empirical mobility exhibit consistent shifts in peak timing and curve shape across SEIR parameter scenarios. In contrast, simulations driven by scale-free mobility exhibit less variation and weaker temporal agreement. These findings underscore the importance of realistic mobility flows for capturing scenario-dependent epidemic dynamics.