Analysis of epidemic spread dynamics using a PDE model and COVID-19 data from Hamilton County OH USA

We study the spatiotemporal dynamics of an epidemic spread using a compartmentalized PDE model. The model is validated using COVID-19 data from Hamilton County, Ohio, USA. The model parameters are estimated using a month of recorded data and then used to forecast the infection spread over the next ten days. The model is able to accurately estimate the key dynamic characteristics of COVID-19 spread in the county. Additionally, a stability analysis indicates that the model is robust to disturbances and perturbations which, for instance, could be used to represent the effects of super spreader events. We also use the modeling framework to analyse and discuss the impact of Non-pharmaceutical interventions (NPIs) for mitigation of infection. Our results suggest that such models can yield useful short and medium term predictive characterization of an epidemic spread in a restricted geographical region and also help formulate effective NPIs for mitigation. The results also signify the importance of further research into the accurate analytical representation of specific NPIs and hence their dampening effects on an infection spread.