Thread by @mahmudme01

We mostly studied US covid cases last year Fall-Winter time. We found that there is a certain indoor wet bulb temp which is more contagious than the others. This calculation is approximately accurate when people are supposed to use heaters or do not use indoor ACs.

However, during cooling time calculation and prediction is far more complicated. Therefore, I checked US COVID cases for weather conditions from March 20th 2021 to March 28th 2021 when US had a plateau or slight increase in cases. My calculation includes 1532 US counties.

I found exactly the same pattern as what I found from last year's case studies. There is an existence of a virus fav indoor climate. Here is my validation argument- you can get a low number of cases for any indoor condition because there may have multiple factors involved in it.

However, cases must go high only for a certain wet bulb temperature. If any data show up in a non-virus fav climate, my argument is invalid. Therefore, you can see 3 outliers in this plot where cases were high in non-virus fav indoor conditions (marked by arrows).

From the 'error bar', you can see that indoor WBT varied significantly during those 9 days periods. Let's look at Chattahoochee County, GA (top arrow). The red rectangle is the weather-analyzed window. 7-day average covid data is shifted 10 days towards left.

4 days were in virus fav condition (black dots) and covid cases went up. Then it was plateaued in non-virus fav cond. Next Matagorda, TX (shown by bottom arrow). You can see a similar pattern. Another dot (next to the same arrow) is the nearby county- Karnes, TX.

It is difficult to locate the exact indoor WBT from this vast number of data analyses. There are uncertainties for different reasons. For example, indoor WBT is calculated considering that people lived indoors at 21C. Virus variant may add uncertainty too. nrel.gov/docs/fy17osti/…

Here, I have shown that indoor WBT modulates COVID cases (in fact all respiratory illnesses). Now, it's time to do some controlled experiment to figure out the exact temp at which people may get sick. It may differ from people's age and height too. researchgate.net/publication/34…

Here is the same plot with 339 counties (only those counties which reported more than 100 cases during that time.

The same plot for those counties whose population density is more than 100 (number of data points 423)