Probability of Rain

[Guy Berger]

The weather.com forecast usually includes an hourly probability of rain, 0%, 10%, 20% etc. My question is whether this probability is hourly only. So for example, if for each hour for the next 3 hrs it says "20% probability of rain" does that imply that the probability for rain sometime over the next three hours is 48.8%? (1 - [0.8*0.8*0.8]) Or is that just short hand for "20% probability of rain for the next few hours"?

Guy

[Dan Gould]

20% rain chance means that 20% of the geographic area under consideration will see rain in a given time period. It is not cumulative.

Edited July 27, 2005 by Dan Gould

[Michael Fitzgerald]

What? That makes no sense. So the chance of rain is 100% for some 20% of the geographic area?

Let's say the geographic area is Pennsylvania. Let's say the window is 3 hours. It will definitely rain somewhere in Pennsylvania during those 3 hours and the total of all the places where it rains will add up to 20% of the state?

Can't it just not rain at all anywhere in Pennsylvania during the entire 3 hours - and "oops, that was the 80%"?

Mike

[Dan Gould]

What? That makes no sense. So the chance of rain is 100% for some 20% of the geographic area?

←

No, because there is no guarantee that rain will fall. We're talking about statistical probability, not guarantees.

Let's say the geographic area is Pennsylvania. Let's say the window is 3 hours. It will definitely rain somewhere in Pennsylvania during those 3 hours and the total of all the places where it rains will add up to 20% of the state?

Can't it just not rain at all anywhere in Pennsylvania during the entire 3 hours - and "oops, that was the 80%"?

←

It will not "definitely" rain somewhere. The coverage of potential rain will be 20%.

Yes there can be no rain in the given time frame, because we are talking about probability NOT absolutes.

Here's another way of putting it:

Under the present atmospheric conditions, it is likely that no more than 20% of the coverage area will see rainfall. In other words, it is probably that your area will not see rain, but if it does, don't say we didn't tell you so.

If you haven't figured it out by now, rainfall probabilities are not an exact science, or have you never had a picnic ruined by rain?

[Michael Fitzgerald]

I guess I'm confused by what the 20% is - I would have thought that would be me flipping a coin 100 times and for only 20 of those it would have been raining, not anything to do with geographic area.

How's this -

http://www.wrh.noaa.gov/byz/wxtalk/talk16.php

That seems to be more in line with what I'm thinking.

Mike

[Dan Gould]

Well, the way I was taught (admittedly I was a student for only a brief time) by the Meteorology Department of Florida State was in terms of percent of a geographical region. Maybe definitions have changed since the mid 80s.

[Michael Fitzgerald]

Hey, you know what they say: a little bit of knowledge.....is more than I have on the subject.

Mike

[.:.impossible]

I live in an area where there is a 30% chance of rain every day during the summer. It doesn't usually rain everyday, but there's always that chance. Thunderstorms bloom up on the doppler out of nowhere. Usually not in my backyard however.

In the Virgin Islands, a storm will pass through every day pretty much without fail from April through Hurricane season, probably year 'round, I don't know. Forecast: chance of thunderstorms 100%. Easy.

[JSngry]

20% chance of rain means that they're not predicting rain but are covering their ass anyway, just in case they're wrong.

[Dave James]

Any weather forecast anywhere by anyone is not meant to be taken seriously. If there is a profession (and I use this term in its loosest possible context) that is more wrong more often than this one, enlighten me as to what it might be. One of the biggest ruses ever perpetrated on the American public is the so-called 10 day forecast. These people don't know what it's going to do tomorrow much less 10 days from now. If you sat outside in a shed with your big toe in a bell jar, you could peg the weather as well or better than the thimble brains who do this for a living.

Up overcast and out.

[Michael Fitzgerald]

I'm not sure which is the prettier picture (big toe in bell jar or brain in thimble).

Mike

[Man with the Golden Arm]

Triumph's forecast!

a quarter of the way down...

[Dan Gould]

Any weather forecast anywhere by anyone is not meant to be taken seriously.  If there is a profession (and I use this term in its loosest possible context) that is more wrong more often than this one, enlighten me as to what it might be.  One of the biggest ruses ever perpetrated on the American public is the so-called 10 day forecast.  These people don't know what it's going to do tomorrow much less 10 days from now.  If you sat outside in a shed with your big toe in a bell jar, you could peg the weather as well or better than the thimble brains who do this for a living. 

Up overcast and out.

←

Yes, 10 day forecasts are ridiculous and a ruse. But meteorology is hardly a profession to be laughed at. At three and 5 days, forecast accuracy is extremely high, and as a former meteorology major, I find your comments extremely offensive.

So, Dave, do you rely on your own toe in a bell jar to decide whether to bring an umbrella with you to work or do you still actually listen to the "thimble brains" who populate this "so-called profession"??

[Dan Gould]

Any weather forecast anywhere by anyone is not meant to be taken seriously.  If there is a profession (and I use this term in its loosest possible context) that is more wrong more often than this one, enlighten me as to what it might be. 

←

Try economics.

[Dave James]

I live in Portland, Oregon. If I understand correctly, one of the more difficult places in the country to predict the weather. The most accurate and reliable forecast I'm aware of in this area comes in the form of a "beacon" that sits atop a building downtown called The Standard Plaza. If it's white, it will be cooler. If it's green, no change. If it's red, warmer. If any of the colors flash, you can expect some precipitation. Another technique I practice - hang onto your hat - is to go outside in the morning and look up at the sky. Unbelievably accurate.

Of course I listen to the weather forecast, but you really don't have much choice if you're the kind of news junkie I am. I just don't pay much attention. For a group of people who are right maybe half the time, they take themselves way, way too seriously.

With regard to ecomomics and accuracy I suppose there may be some validity in the comparison although I rarely get caught in the rain without an umbrella if someone says the GNP is slipping. Point being, a weather forecast is immediate and has immediate impact. Economics exists on a wholly different level.

If my comments were offensive, that was not my intent, although that characterization strikes me as a bit overreactive given the marginal seriousness my original statement. I guess what really bothers me about weathermen or anyone else for that matter is someone who acts like they know what's going on when they really don't. Come to think of it, George Bush would make a helluva weatherman.

Up over and out.