Evaluating COVID-19 Mitigation Strategies in Schools with a Spatially-Explicit Agent-Based Model of Infection Dynamics

This project will develop and refine a simulation modeling system to help schools and school districts evaluate the effects of interventions to control COVID-19 infections among students, focusing first on schools in San Diego County, CA. The goal is to help individual schools choose the most effective strategies for their unique circumstances, particularly each school's population, classroom layouts and ventilation, neighborhood infection rates, and school bus commutes. The interactive model will allow decision makers at the school and district levels to evaluate the potential impacts of both pharmaceutical interventions, like vaccination and testing, and non-pharmaceutical safety measures including social distancing, mask-wearing, limiting interactions between different student cohorts, decreasing classroom occupancy, and improving ventilation. Finding the best combination of such measures is central to safe school re-opening, which is critical for improving child education and development and allowing parents to return to work. As school-aged children, many of whom are not vaccinated, are disproportionately affected by new and more dangerous COVID-19 variants, this simulation service will provide much-needed insights to our partners among decision-makers at the school and district level in San Diego County and elsewhere in the nation.

The school infection model simulates aerosol and droplet transmission as students, teachers, and staff (modeled as human agents) interact during typical school day activities, including classroom instruction, cafeteria lunch, recess, and traveling on a school bus. The likelihood that a healthy agent is exposed to the virus increases as they come near infected, asymptomatic agents or spend significant time in poorly ventilated spaces. The model considers spatial information about room layouts and ventilation, and simulates school day schedules down to 5-minute intervals. Building on a successful prototype for elementary schools, this project will extend the model to middle and high schools and to neighborhoods with different socio-economic and demographic characteristics. Additionally, it will enable simulation of infection dynamics for athletics activities and analysis of infection patterns under different COVID-19 variants. Making the model accessible via a science gateway will let school administrators, researchers, teachers, and students run simulations on a supercomputer, then visualize and compare model outputs under different scenarios. As in the prototype phase, the project will engage undergraduate students in all aspects of model development, and in presenting results to public health and education experts and to the general public - an essential component of their training as data scientists working on societally-impactful topics like education and health equity.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.