Potential improvements to the immunity layer

[emmamendelsohn]

We have a sparse seroprevalence layer in space and time from Clark et al. (shared via Mindy)

As an immunity layer, we could link seroprevalence data versus distance in time and space from outbreaks. Make a GAM where each country is linked in space via Markov chain random field. Baseline model could just be a spline over space and time on the seroprevalence data.

This will be after ARC August 24 deliverable.