A kernel-modulated SIR model for Covid-19 contagious spread from county to continent

Area of Science:

• Epidemiology
• Complex Systems Science
• Statistical Physics

Background:

• COVID-19 spread exhibited complex tempo-spatial patterns influenced by individual and societal decisions.
• Early outbreaks featured distinct infection hotspots, which later evolved towards more uniform spatial distribution.
• Understanding these dynamics across scales is crucial for effective public health strategies.

Purpose of the Study:

• To analyze the spatial evolution of COVID-19 infections in the United States across multiple scales.
• To develop a novel modeling approach for characterizing disease spread in multifractal populations.
• To explain the occurrence of multiple epidemic waves despite nonpharmaceutical interventions.

Main Methods:

• Analysis of COVID-19 case data across scales from county to continental levels (10 to 2,600 km).
• Application of multifractal scaling analysis to understand spatial distribution patterns.
• Development and implementation of a kernel-modulated susceptible-infectious-recovered (SIR) model.

Main Results:

• COVID-19 spatial distribution in the US followed multifractal scaling, evolving over time.
• Spatial correlation increased rapidly early in the outbreak, then slowed to approach population correlation.
• The kernel-modulated SIR model successfully reproduced multiphase epidemics, explaining waves linked to population flow changes.

Conclusions:

• The study offers an original interpretation of COVID-19 spread dynamics, emphasizing multifractal characteristics.
• A pragmatic kernel-modulated approach effectively captures spatial intermittency and disordered patterns.
• The model explains recurring epidemic waves due to susceptible population movement and policy changes.