2 ms·
Note that if the climate at a place isn't changing we'd expect new records for a given thing to become less frequent over time. For example suppose you had some
by tzs 2y ago
Note that if the climate at a place isn't changing we'd expect new records for a given thing to become less frequent over time. For example suppose you had something where it follows a normal distribution with a mean of 100 and a standard deviation of 5. I generated 10 million random numbers with that distribution and noted whenever I got a new record high.
The intervals between records were 1, 1, 2, 1, 5, 36, 219, 999, 2558, 21780, 397, 50360, 73288, 200951, 652234, 954240, 547715, 3783984, and 1641056.
- dmurray 2y agoThis is both completely true and completely nonsense in the context. If the numbers are independent, no matter what the distribution, the odds of the nth measurement being a record setter is 1/n. And as above, if you've measured 100 Novembers, the odds of the 100th being the record wettest are 1/100.