Uncommon Cold Doesn’t Break the Rules

I’ve done a bit of work on COVID vs latitude.   The results are interesting to me, colds are caused by the same family of virus as COVID 19. I have been very curious if, like the common cold, this coronavirus would be impacted by climate.

From Johns Hopkins university, they provide a map of COVID deaths across the united states.    The cities stand out but it is impossible to say that the North or South did worse.  In the ShenZhen area of China, the people are very densely packed, yet the virus didn’t impact them as severely as New York.  I’ve seen their hospital system personally and I can promise you that, that is NOT what solved it.  But the city is just at the Tropic of Cancer and is generally warm and humid.  Not a great place to catch a cold.

My theory was that if I were to only look at counties with enough population density and sum by latitude, I might see a pattern to Deaths from COVID.  What I found was rather striking.  Considering that Michigan had some of the strictest lockdown rules and suffered terrible stats compared to Texas or Florida,  I have been wondering why.   In the map below, you can see Iowan cities didn’t do well, neither did Minneapolis.  Quite a few counties carry so little population that they would have effective quarrantine so I needed to remove them from the stats.  I downloaded census data and sq mile data put it all together over about 9 hours of work. I used deaths instead of cases because cases are dependent on testing availability and I would expect that to add more noise to the data.  All that said, I think I will let the graphs do the rest of the talking.

 

That was a little surprising to me.  Such a strong grouping at the same latitude as southern New York to Just north of Michigans lower peninsula.  Just to confirm what I was seeing, I ran the population density per latitude below.

I’m thinking that this must be caused by behavior.  I know that we stay in during the cold months and head into the sunlight when it is warm.   It’s not warm in Michigan and New York until right about now.   This could be good news in the coming days.

I did run a bunch of curve fits on the data and found the 95% CI outside of the zero range.  Also did some multivariate regression and found that Deaths were explained by latitude, population and population density outside of the CI again.