Skip to content

Raw data: Heat-related deaths in the US

This is from a research letter in JAMA today:

Average temperatures in the US have gone up 2°F since 2000. That might not seem like much, but when the average increases a little the tail of the bell curve increases a lot. In the past decade, that small increase in average temps has caused the number of extreme heat waves to increase by upwards of 3x. Thus the very large increase in heat-related deaths.

7 thoughts on “Raw data: Heat-related deaths in the US

    1. skeptonomist

      Just ignore the trend lines - Kevin puts these in for artistic reasons. Obviously just looking at the blue lines almost all increase came after 2015.

  1. skeptonomist

    "when the average increases a little the tail of the bell curve increases a lot."

    Not if the shape is the same. If you want to say something about the tails, talk about the actual data - how did the standard deviation change? How about the frequency of really hot days?

    But two degrees does sound like a lot and the response of humans to temperature is not necessarily linear.

    1. FrankM

      If the shape of the curve stays the same but the mean changes, then it's the probability of events at the tails that changes most dramatically.

    2. TheMelancholyDonkey

      Not if the shape is the same.

      Yes, if the shape is the same.

      Probabilities in the tails change pretty rapidly with a change in the mean. For instance, with a normal distribution with a standard deviation of 1, the probability of an event at least 3.8 SDs from the mean is more than twice the probability of an event at least 4.0 SDs away. The probability of that event at least 4.0 SDs from the mean is more than three times the probability of an event 4.2 SDs away.

Comments are closed.