UROP Openings

Have a UROP opening you would like to submit?

Please fill out the form.

Submit your UROP opening

Looking Past the First Wave of the COVID-19 Pandemic: Metapopulation Modeling to Design Strategies for Lockdown Scale-back


Term:

Summer

Department:

1: Civil and Environmental Engineering

Faculty Supervisor:

Ruben Juanes

Faculty email:

juanes@mit.edu

Apply by:

May 7, 2020

Contact:

Christos Nicolaides: chrisnic@mit.edu

Project Description

Epidemiological modeling work on COVID-19 has, up to now, focused on the initial spread of the disease, for example, evaluating the impact of travel restrictions within China and worlwide on the progression of the pandemic. The pressing question, currently, is a different one: how will governments lift the wide range of lockdowns to reactivate their economies while preventing a renewed explosive spread of the disease? We address this question abstracting the different types of intervention into three categories: (1) long-range travel through the air transportation network (Nicolaides et al., PLoS ONE 2012); (2) regional or city-scale social distancing measures, such as business and school closures; and (3) individual hygiene measures, such as the use of face masks and hand washing (Nicolaides et al., Risk Analysis 2020). We will incorporate these three levels of intervention into a global metapopulation disease transmission model. This metapopulation network approach will integrate real-world data at various spatial scales, accounting for the coupled evolution of disease spreading and human mobility. We will define subpopulations associated with major transportation hubs (i.e., airports). We will capture the infection dynamics within the subpopulations as a result of human contact and disease awareness, and between subpopulations as a result of human travel. We have real-world data to calibrate, constrain and initialize the model, and to use it in forecast mode to assess the merits of different intervention scenarios. Using data of confirmed cases and deaths around the world, we will first calibrate the model to simulate a wave of contagion that reproduces a global pandemic state. The emphasis is not on replicating the detailed statistics that we see at a country level or regional level but, rather, on the statistical heterogeneity of disease prevalence that we observe globally. Using that state as initial condition, we will evaluate different strategies for scaling back the lockdown measures at all three spatial scales: long-range travel, city-scale mobility, and personal hygiene. Each strategy will result in one or more spikes of COVID-19 contagion, and we will be able to contrast them in terms of death rate, timing, duration, and robustness.

Pre-requisites

Expertise in computer programming. Familiarity with modeling. Ideally, familiarity with modeling of network processes.