SEVA: An externally driven framework for reproducing COVID-19 mortality waves without transmission feedback

Abstract Understanding the mechanisms underlying epidemic waveform structure remains a central problem in infectious disease epidemiology. During the first COVID-19 wave, epidemic time series across regions displayed markedly different temporal patterns, including sharply peaked curves with rapid decline and more prolonged plateau-like trajectories. Here we present a parsimonious dynamical framework for describing epidemic waveforms at the level of observable endpoints. In the SEVA (Seasonal/Environmental Viral Activity) formulation, epidemic incidence arises from the interaction between a time-varying activity function, interpreted as a population-level hazard, and depletion of a finite vulnerable population. With activation timing and steepness held constant, variation in activity intensity governs the resulting waveform dynamics. The framework is applied to hospitalization and mortality data from European countries and U.S. states during spring 2020. Using a common activation profile, the model reproduces both sharply peaked and plateau-like epidemic regimes across regions without modification of the underlying dynamical structure. While interpersonal transmission undoubtedly occurs, the observed epidemic waveforms are not readily explained by transmission feedback alone. A key empirical observation is that normalized epidemic trajectories exhibit closely similar temporal structure across regions with widely different mortality burdens. Within the proposed framework, this behaviour can be understood as arising from the interaction between a shared temporal activation profile and region-specific variation in activity intensity. These results show that diverse epidemic waveform regimes can be captured within a simple dynamical framework operating at the level of observable endpoints. The SEVA framework therefore provides a testable hypothesis for the dynamical origin of epidemic waveforms that can be evaluated across different pathogens and epidemiological settings.