Individual dynamics and local heterogeneity provide a microscopic view of the epidemic spreading

Abstract The COVID-19 pandemic had caused severe global disasters, highlighting the importance of understanding the details and trends of epidemic transmission to introduce efficient intervention measures. While the widely used deterministic compartmental models have qualitatively presented continuous "analytical" insights and captured some transmission features, their treatment usually lacks spatiotemporal variation. Here, we propose a stochastic individual dynamical (SID) model to mimic the random and heterogeneous nature of epidemic propagation. This SID model provides a unifying framework for representing the spatiotemporal variations of epidemic development by tracking the motion of each individual. Using this model, we reproduce the infection curves for COVID-19 cases in different areas globally and find the local dynamics and heterogeneity at the individual level that affect the disease outbreak. The macroscopic trend of virus spreading is clearly illustrated from the microscopic perspective, enabling a quantitative assessment of different interventions. Seemingly, this model is also applicable to studying stochastic processes at the "meter-scale", e.g., human society's collective dynamics.