June 22-July 31, 2020
This was the week where the rubber hit the road. Having established five “umbrella groups,” each focused on a particular theme, and grouped fifteen projects under the various umbrellas, it was time to get to work and see how to apply mathematics, simulations, and data assimilation to get a better understanding of Covid-19.
The daily schedule had a few fixed points: All-Hands meetings at 11 am and 5 pm, tai-chi (optional) at 3 pm, and tutorials during the lunch breaks (see below). Most of the time was spent on individual and group discussions, reading, developing mathematical models, programming them in Matlab or Python, and narrowing down the scope of the projects.
Five Umbrella Groups and their Subgroups
Diseases and the Environment
- Air Quality
Impacts of Behavior
- Modeling human behavior using game theory
- Adaptive network structures for social interactions
Incorporation of Data
- Multi-scale DA
- Blood types and COVID
- Topological epidemiology
- Optimal control, testing
- Network topologies for testing strategies
- Optimizing pop-up testing sites
- Uninsured/insured populations
- Economics and disease interactions
- Age-structured multi-population models
- Multi-population models
The learning process, which had occupied most of Weeks 1-3, was continued on a smaller scale with two special lectures. On Wednesday, Steffen Eikenberry (Arizona State U) gave a talk on “To Mask or Not to Mask,” and on Thursday, Linda Allen (UC Boulder) returned with a tutorial talk on “Branching Processes.”
Steffen’s talk addressed the effectiveness of facemasks to curtail the spread of a pandemic. His talk was based on a recent paper published online in Infectious Disease Modeling(April 21, 2020). Apparently, the topic was already discussed in an early paper by A.J. Jessup, published in Scientific American SupplementNo. 143 (1878) but remains controversial and is sometimes painted as a new symbol of tyranny. The talk included details about the Manchurian pneumonic plague (1910-’11), the Spanish influenza pandemic (1918-‘19), the SARS epidemic (2002-’04), and the current Covid-19 pandemic. Steffen emphasized that the effectiveness of wearing masks can be studied effectively with an extended version of the SEIR model. The bottom line was clear: universal application of mask use is critical to curtil the spread of a pandemic.
Linda’s talk was a follow-up on her talk during Week 2. This time, she focused on estimating the probability of an outbreak of an epidemic using branching processes, a new topic for most of the students.
This week’s program featured three tutorials given by the mentors during lunch breaks. Topics covered were Github (for code sharing), Zotero (for reference material), and fundamentals of data assimilation.