Accurately Modeling Coronavirus Spread in Segments

Self-organized diffusion models have been recently used to simulate the dynamic spread of the COVID-19 pandemic. These models contain a critical point where the duration of the pandemic is maximized—a condition that policymakers and the world aim to avoid.

Originally studied with sandpiles, self-organized diffusion models set specific diffusion rules, then leave the system to organize itself. Sand grains are found at the nodes of a lattice. The nodes offer and accept grains from other connected nodes.

After setting the initial density of grains, the model outputs the nodes that are capable of diffusing grains and the evolution of the active node percentage.

Contoyiannis et al. continue to build on this process with the Method of Parallel Trajectories, which simulates the pandemic in segments of approximately 20 days.

“When the curve of the model produced by a specific initial density ceases to closely monitor the epidemiological curve, then another one is activated with a different initial density that closely monitors the next part of the epidemiological data,” said author Stavros Stavrinides.

For example, if a five month data curve arises and it is divided into six parts, there are six parallel model trajectories that all have the same dynamics, but come from different initial conditions activated in successive time intervals.

The team tested the model on data from third wave pandemic dynamics in Greece and Italy. Whenever restrictive measures were loosened or removed, they found the epidemiological indicators worsened considerably.

The model allows for short-term infection forecasting and close monitoring of the pandemic in comparison to the critical point. The researchers aim to consider vaccinations in their model as a future improvement.

Source: “Application of the method of parallel trajectories on modelling the dynamics of COVID-19 third wave,” by Y. Contoyiannis, S. G. Stavrinides, M. P. Hanias, M. Kampitakis, P. Papadopoulos, R. Picos, S. M. Potirakis, and E. Kosmidis, Chaos (2022). The article can be accessed at https://doi.org/10.1063/5.0075987 .