Thread by @jnordvig

I will do what everybody knows is very dangerous on twitter: comment on Sweden...

Cases have accelerated sharply over the last few weeks, from 200/day to 700/day (could have broader implications, with weather possibly a key variable = THREAD).

I am not trying to be alarmist, very few people seem to be getting seriously ill, as evidenced by low ICU admissions (although there is a slight uptick, and obviously lags at play)

But it raises the question. WHY? (are cases spiking)

Sweden has been 'open' for a while (that is the reason everybody is so focused on Sweden's alternative model)

Relatedly, mobility indicators have been pretty stable over the last few months.

Could it be the weather?

In the US the virus is moving to the North (as indicated by the hotspot map, which have few clusters in the warmer south)

https://twitter.com/ExanteData/status/1317626845106475014

There are other possibilities (social distancing fatigue, return to work/school after summer, hang-over from August tourism travel). But the weather is a factor too, and it is worth thinking pretty hard about how Rt (all else equal) will be impacted by temperature.

I have seen very little academic/epidemiology commenting on weather effects, except in a generic modelling sense. Perhaps it is under-researched still, and could entail a surprise element as weather gets colder and colder in the Northern Hemisphere.

I will leave it at that. END