PANDEMIC: To Simulate Infection Spread in a Complex Environment
Abstract
PANDEMIC is a new, time-dependent simulation model treating the infectious spread of a disease in a complex, geographically varied region containing millions of people. In PANDEMIC, each person in the region moves from place to place and interacts with others they meet while commuting to work, going to meetings (i.e. church and other events), to shop or school, and returning to or remaining at home. This model takes advantage of parallel computing using an NRL-developed algorithm called the Monotonic Lagrangian Grid (MLG) to keep track of who is close to whom, as the people follow their usual routines. This model allows what if tests of different personal and societal public health behaviors and various mitigation strategies to compare their impact and effectiveness. The goal of this research is to predict in greater detail the dynamics of mitigating the current covid-19 pandemic while developing more efficient algorithms and computational techniques using new computer hardware designs. PANDEMIC simulations are presented predicting the three covid-19 infection surges observed through January 2021. Different levels of testing to mitigate the catastrophic third surge are compared. Another prototype study, illustrating how the model can be used for hypothesis testing, is also reported. It evaluates the impact of largely unmasked super-spreader events showing how each such gathering ultimately increases the number of covid-19 related deaths by hundreds.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 17, 2021
- Accession Number
- AD1148238
Entities
People
- David R. Boris
- Jay Boris
- Keith Obenschain
Organizations
- United States Naval Research Laboratory