Contagion processes, be it the unfolding of an infectious disease epidemic or the dissemination of misinformation in a population, are multiscale phenomena that spread through individuals' interactions. At the finest resolution, these interactions may take the form of a network interconnecting each individual, rich with heterogeneities, and may vary in time. On a larger scale, entire subpopulations—which might range from a small city to a country—influence each other because of human mobility, notably shaped by the global air travel network.

In this presentation, we will start at the individual level and discuss the modeling of contagion on networks, emphasizing the important roles of heterogeneity, group structure, and modeling assumptions in shaping the properties of spreading processes. In the second part, we will move to a population-level description of contagions, allowing us to model spreading processes on a global scale. Achieving realistic modeling at this scale, however, imposes a significant computational burden. To circumvent this issue, we introduce a novel metapopulation framework that leverages probability generating functions. This approach notably enables us to rigorously model and quantify the performance of a global wastewater surveillance system at airports for detecting emerging pathogens. Finally, we will discuss future directions for extending this computational platform, how to bridge the seemingly disconnected modeling scales, and how this research aligns with a broader vision to improve population health and better prepare for the next pandemic.