Probability of rain

Sorry Joyce, but you are wrong. Cecil’s answer is close. The probability of rainfall is based on current conditions and past performance. It is the chance that for a randomly chosen point in the area of prediction it will rain there at some time during the period of prediction.

The reference is to What does it mean when “X percent chance of rain” is predicted? - The Straight Dope, where Joyce K. was quoting as writing:

I’m pretty sure Joyce was joking.

However, I had a question about anothing statement in this 1986 post, where Cecil says, “The weather service has been calculating precipitation probabilities for as long as anybody can remember, but it’s only been in the last decade or two that the nation’s cadre of broadcast weather beings has deigned to convey this to the masses.” When did this actually start? I’m pretty sure it was being done in the early 1970s. I recall a cartoon in that period from a farming magazine, in which a cow was reluctant to leave a barn, seemingly trying to listen to the radio there, and the farmer said, “Go on, [cow’s name]! You don’t need to know the chances of measurable precipitation!”

I’ve also heard the interpretation that a 30% chance of rain means 30% of a region will get rain, which seems to hold true when taking a road trip because you will hit areas of rain whereas at home it stays dry.

That really only works if they know with 100% certainty what portion of the viewing/listening area will get precipitation. Which they don’t. So you’d still need to account for the probabilities of 10% of the area getting rain, or 50% of the area, or whatever.

For small and large percentages, this amounts to nearly the same thing (as % for a random location). But it doesn’t handle the case where there’s a 50% probability that all the area will get rained on (say, there’s a big cloud coming, and if the temperature drops enough by the time it gets here, it’ll rain everywhere, but if the temp doesn’t drop or the front stops moving, it won’t.)

The random-point probability is mathematically more well-defined, and still eady for a listener to interpret correctly: “Gee, I have about x % chance of being rained on, at some point today.”

I’m just glad this column repeated, because I hear all the time people repeating the claim that x% chance of rain means that in a given area x% will get rain. That irks me to no end, so I’m glad I finally have Cecil’s column to reassure me these people are [del]idiots[/del] misinformed.

Being a Pacific Northwest native myself, I’m sure she was joking.

OTOH, it does seem to be pretty accurate. :wink:

Permutations, probability, averages, etc are still all guesses. It just seems counter-intuitive when I’m driving in pouring rain, visibility barely beyond my hood, and the radio tells me there’s a 20% chance of rain today.

Here is what the National Weather Service says about Probability of Precipitation statements:

FAQ - What is the Meaning of PoP"?

Statistics aren’t real time indicators. They are predictions.

But how do they determine the area that is expected to get precipitation? Seems odd to me.

I think you’re confusing statistics with probability.

My guess is that most models are finite-element grids, and the algorithm walks the grid and asks itself (for example) “for this 0.001% of the Greater Metro area, what’s today’s probability?” and does that for all 100,000 elements.

The broadcast result is then computed as the average over those 10K elements.