[tertius]

What is the probability of a 3 days in a year breaking high temps? I'm not sure how to read the daily high numbers here unless I know what number is bad/good.

How likely is it in any given year in the last 100 years that there wasn't a day or 3 breaking previous highs?

That seems like weather vs. climate to me intuitively, but I just don't know.

[adhesive_wombat]

Impossible to say without knowing the variance (and actually the distribution in general) of the data. But it would be nice to see a trend of time between records: if temperatures are not changing, we'd see the intervals increasing (setting records from a random distribution gets rarer and rarer). But if they're constant or increasing, that's a sign temperatures must be increasing on average.

However, that said, 9 of 10 ten of the hottest years on record happened within the last 10 years[1] (or if NOAA is not trustworthy to you, all 10 of the hottest UK years on record are since 2002 [2]), which would be extremely unlikely to be random chance, even if temperatures were completely random.

It would take much deeper statistical analysis then I can do to tell you the sigmas of those events, though.

[1] https://www.noaa.gov/news/2021-was-worlds-6th-warmest-year-o...

[2] https://www.metoffice.gov.uk/about-us/press-office/news/weat...

[titzer]

Here's [1] an animation of real data from NASA.

The probability question is maybe just a math exercise at this point.

https://www.youtube.com/watch?v=Yy0mN_t_3x4