search

ncase / covid-19

COVID-19 Futures, Explained With Playable Simulations
Creative Commons Zero v1.0 Universal
233 stars 106 forks source link

Fix Celsius-Fahrenheit conversions #18

Closed gereeter closed 4 years ago

gereeter commented 4 years ago

I'm not entirely sure where the existing numbers come from; besides the conversion rate being 1.8, not 2.2 (1kg = 2.2lb, perhaps?), 2.2*15 = 33. The citation uses Celsius, so I kept those numbers the same.

gereeter commented 4 years ago

Hm, using the 2019 average monthly temperatures in Central Park (NWS), the difference between the hottest month and the coldest month is 47.1°F or 26.2°C. Meanwhile the simulations use a value of 0.4, citing 15°C0.0225, but `150.0225 = 0.3375`.

ncase commented 4 years ago

Frick, thanks for catching this! How embarrassing. I made, like, three different mistakes simultaneously:

1) Yup, I confused the metric/Imperial conversion for pounds vs Fahrenheit 2) To convert "15°C" to Fahrenheit, I just typed "15°C in F" on Google and it said 59°F, but I forgot that it's converting absolute degrees, not difference 3) The simulation said 15*0.0225 because I screwed up the calculation twice. One, yeah it should be 0.3375 not 0.4. And two, the "Reduction of R by 0.0225" number comes from Wang et al, but I forgot to divide by their estimate of R = 1.83.

Huge thanks again! I also linked to the weather.gov link you posted as a source for the summer-winter difference in NYC, thanks for finding a more solid source for that!