# 40% chance of precipitation: how to parse this? **URL:** \ **Category:** Factual Questions\ **Created:** [May 14, 2011, 4:45am UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876 "2011-05-14T04:45:46Z")\ **Posts on this page:** 16\ **Page:** 1
**Author:** ![VarlosZ](https://avatars.discourse-cdn.com/v4/letter/v/d78d45/32.png) [@VarlosZ](https://boards.straightdope.com/u/VarlosZ)\ **Post date:** [May 14, 2011, 4:45am UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876/1 "2011-05-14T04:45:46Z")
So I’m thinking of going to Atlantic City tomorrow, but I want to check [the forecast](http://www.wunderground.com/cgi-bin/findweather/getForecast?query=Atlantic%20City,%20New%20Jersey). As you can see, they give a 40% chance of showers. If you break it down [hourly](http://www.wunderground.com/cgi-bin/findweather/getForecast?query=Atlantic%20City,%20New%20Jersey&hourly=1&yday=133&weekday=Saturday) (actually, into 3-hour intervals), it lists a 40% chance of showers at 11 AM, 2 PM, 5 PM, and 8 PM. Now, does this mean that, within that 9 hour time span, there is a 40% chance of measurable precipitation occurring at some time? Or, does it mean that there’s a 40% chance it will be raining at 11 AM, **and** a 40% chance it will be raining at 2 PM, etc.? I’m guessing it’s the former (because if it’s the latter then the general forecast of “40% chance of showers” is superficially misleading), but I wanted to be sure. Thanks. ---
**Author:** ![running\_coach](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/running_coach/32/15836_2.png) [@running\_coach](https://boards.straightdope.com/u/running_coach)\ **Post date:** [May 14, 2011, 4:47am UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876/2 "2011-05-14T04:47:58Z")
[Here’s a good explanation.](http://www.srh.noaa.gov/ffc/?n=pop) ---
**Author:** ![VarlosZ](https://avatars.discourse-cdn.com/v4/letter/v/d78d45/32.png) [@VarlosZ](https://boards.straightdope.com/u/VarlosZ)\ **Post date:** [May 14, 2011, 4:50am UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876/3 "2011-05-14T04:50:28Z")
I actually read that before posting. It does a good job of discussing the affect of variable rainfall over an _area_, but, unless I’m being dense, it doesn’t seem to address rainfall over _time_, which is what I’m wondering about. ---
**Author:** ![Jman](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/jman/32/4416_2.png) [@Jman](https://boards.straightdope.com/u/Jman)\ **Post date:** [May 14, 2011, 1:43pm UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876/4 "2011-05-14T13:43:52Z")
My understanding is that it represents historical data. That is, the projected conditions at 2PM (temperature, humidity, pressure, winds, etc) have resulted in rain (when looking at historical conditions that are similar) 40% if the time. (or as a combination of frequency and area as described in the previous link) So, basically, in days like tomorrow, it has rained 40% of the time. ---
**Author:** ![Arjuna34](https://avatars.discourse-cdn.com/v4/letter/a/eb8c5e/32.png) [@Arjuna34](https://boards.straightdope.com/u/Arjuna34)\ **Post date:** [May 14, 2011, 1:50pm UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876/5 "2011-05-14T13:50:22Z")
Cecil on the subject: [Whatdoes it mean when “X percent chance of rain” is predicted?](http://www.straightdope.com/columns/read/854/what-does-it-mean-when-x-percent-chance-of-rain-is-predicted) Although he doesn’t mention the time aspect. I had interpreted it as AND, i.e. 40% chance at 11am, and a separate, independent 40% chance at 2pm. But, as you point out, that doesn’t agree with a generic overall “40% chance” for the day. ---
**Author:** ![ITR\_champion](https://avatars.discourse-cdn.com/v4/letter/i/c67d28/32.png) [@ITR\_champion](https://boards.straightdope.com/u/ITR_champion)\ **Post date:** [May 14, 2011, 11:51pm UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876/6 "2011-05-14T23:51:42Z")
The two possibilities listed in the OP are not necessarily contradictory. It could be that the forecasters are looking at two scenarios and they give the first a 60% chance and the second a 40% chance. In the first scenario there’s no rain all day, while in the second scenario it rains continuously all day. While thay may be an oversimplification I think that it captures the basic idea. Hourly reports about rain are not independent events. In other words it’s far more likely that either it will rain at both 5 and 8 or it won’t rain at either 5 or 8 than that it will rain at one time but not the other. ---
**Author:** ![ultrafilter](https://avatars.discourse-cdn.com/v4/letter/u/3d9bf3/32.png) [@ultrafilter](https://boards.straightdope.com/u/ultrafilter)\ **Post date:** [May 14, 2011, 11:59pm UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876/7 "2011-05-14T23:59:16Z")
> [@ITR\_champion](#): > > In other words it’s far more likely that either it will rain at both 5 and 8 or it won’t rain at either 5 or 8 than that it will rain at one time but not the other. This may be true, but it doesn’t follow from dependence alone. All that you can really say without knowing something about how weather works is that the probability of rain at 5 and 8 is not the product of the probability of rain at 5 and the probability of rain at 8. ---
**Author:** ![LSLGuy](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/lslguy/32/5813_2.png) [@LSLGuy](https://boards.straightdope.com/u/LSLGuy)\ **Post date:** [May 15, 2011, 1:55am UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876/8 "2011-05-15T01:55:19Z")
NOAA’s forecasts have a specific effectivity period & hence duration attached. That often gets trimmed by your local weather news-goof or [wunderground.com](http://wunderground.com) before they provide their version to you. Per **runner pat** ’s link to NOAA’s definition, the percentage is a blend of raw likelihood and expected coverage area. Both are measured over the entire effectivity period of the forecast. So you can look at the pubished POP number as the likelihood that at least some rain will fall on wherever you happen to be standng at some moment over the course of the entire forecast period. Someplace like wunderground is free to use their own procedures & definitions of these terms. So best to ask them directly how exactly they define their terms. If they are offering 3-hour buckets with a POP value, I’d bet they’re using a similar logic. Trying to add different time buckets is silly. The important thing to keep in mind is the area coverage factor. If it was going to be spotty rain but gauranteed to fall continuously someplace in the forecast area from midnight until the next midnight, they could still give a 40% POP value for the whole day and also 40% for each of the eight 3-hour buckets. ---
**Author:** ![Rick](https://avatars.discourse-cdn.com/v4/letter/r/9f8e36/32.png) [@Rick](https://boards.straightdope.com/u/Rick)\ **Post date:** [May 15, 2011, 4:14am UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876/9 "2011-05-15T04:14:54Z")
They have 10 weathermen in the department. Four of them say it is going to rain, the other six say they are nuts. There is your percentage. ---
**Author:** ![psychonaut](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/psychonaut/32/4655_2.png) [@psychonaut](https://boards.straightdope.com/u/psychonaut)\ **Post date:** [May 15, 2011, 10:03am UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876/10 "2011-05-15T10:03:29Z")
> [@ITR\_champion](#): > > While thay may be an oversimplification I think that it captures the basic idea. Hourly reports about rain are not independent events. In other words it’s far more likely that either it will rain at both 5 and 8 or it won’t rain at either 5 or 8 than that it will rain at one time but not the other. Funny, I thought it would be the exact opposite. Since there’s only a certain amount of water in the clouds, then if it rains at 5:00, then it’s less likely there will be any water left to rain down at 8:00. Similarly, if the clouds are becoming full of water, and it hasn’t rained by 5:00, then it’s more likely that it will rain at 8:00. Of course, these scenarios ignore the possibility of cloud movement via wind. ---
**Author:** ![UltraVires](https://avatars.discourse-cdn.com/v4/letter/u/ecccb3/32.png) [@UltraVires](https://boards.straightdope.com/u/UltraVires)\ **Post date:** [May 15, 2011, 1:46pm UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876/11 "2011-05-15T13:46:42Z")
> [@Arjuna34](#): > > Although he doesn’t mention the time aspect. I had interpreted it as AND, i.e. 40% chance at 11am, and a separate, independent 40% chance at 2pm. But, as you point out, that doesn’t agree with a generic overall “40% chance” for the day. What about at 12 and 1? Why no separate independent 40% chance there? Actually, why not a 40% chance at 11:00, 11:01, 11:02, 11:03, etc. Then with each of those 40% chances combined, the probability would approach 100% that it would rain at some time during the day. 🙂 But that can’t be right because the same conditions that exist at 11:00 would likely still exist at 11:01, making it silly to say that there is now a new chance like rolling the dice. But where IS the dividing line for changes that make a “new” chance happen? An hour? A day? A week? I only say this because I don’t understand it either. When is the condition satisfied? If I walk outside and feel one single drop of rain, does that count? Under this system, when would a prediction ever be “wrong” as opposed to: “Well, it rained, so that was the 40% happening” or “It didn’t rain, so that was the 60%”? ---
**Author:** ![johnpost](https://avatars.discourse-cdn.com/v4/letter/j/f17d59/32.png) [@johnpost](https://boards.straightdope.com/u/johnpost)\ **Post date:** [May 15, 2011, 2:26pm UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876/12 "2011-05-15T14:26:53Z")
it can rain on one side of the road and not the other. looking at a network of monitors there are times when a station will have one inch (sometimes 2 and more) of rain and a station 1 mile zero rain. ---
**Author:** ![postcards](https://avatars.discourse-cdn.com/v4/letter/p/8e8cbc/32.png) [@postcards](https://boards.straightdope.com/u/postcards)\ **Post date:** [May 15, 2011, 3:34pm UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876/13 "2011-05-15T15:34:20Z")
It might rain. Have an umbrella handy. ---
**Author:** ![ultrafilter](https://avatars.discourse-cdn.com/v4/letter/u/3d9bf3/32.png) [@ultrafilter](https://boards.straightdope.com/u/ultrafilter)\ **Post date:** [May 15, 2011, 5:46pm UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876/14 "2011-05-15T17:46:42Z")
> [@jtgain](#): > > Under this system, when would a prediction ever be “wrong” as opposed to: “Well, it rained, so that was the 40% happening” or “It didn’t rain, so that was the 60%”? Individual predictions can’t be verified, but we can track the overall accuracy of a forecaster by looking at the proportion of the time that rain happens when that they predicted a certain probability. If that probability is very different from the proportion, they don’t look so good. [This paper](http://docs.lib.noaa.gov/rescue/mwr/078/mwr-078-01-0001.pdf) has more details. ---
**Author:** ![BrotherCadfael](https://avatars.discourse-cdn.com/v4/letter/b/977dab/32.png) [@BrotherCadfael](https://boards.straightdope.com/u/BrotherCadfael)\ **Post date:** [May 15, 2011, 7:52pm UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876/15 "2011-05-15T19:52:00Z")
Of ten days just like this one, you can reasonably expect rain on four of them. ---
**Author:** ![barbitu8](https://avatars.discourse-cdn.com/v4/letter/b/839c29/32.png) [@barbitu8](https://boards.straightdope.com/u/barbitu8)\ **Post date:** [May 15, 2011, 8:19pm UTC](https://boards.straightdope.com/t/40-chance-of-precipitation-how-to-parse-this/581876/16 "2011-05-15T20:19:27Z")
> [@psychonaut](#): > > Funny, I thought it would be the exact opposite. Since there’s only a certain amount of water in the clouds, then if it rains at 5:00, then it’s less likely there will be any water left to rain down at 8:00. Similarly, if the clouds are becoming full of water, and it hasn’t rained by 5:00, then it’s more likely that it will rain at 8:00. Of course, these scenarios ignore the possibility of cloud movement via wind. I assume from this that you have never experienced an all-day rain day. Rain depends upon the moisture in the atmosphere and convection (turbulence). If there’s an occluded or stationary front just to the west, this can result in an unstable atmosphere (much convection). Combine this with high humidity, it can rain at any time and rain frequently. This scenario applies for showers and thunderstorms. However, any approaching cold front can cause uplifting ahead of it because cold air is denser than warmer air. The cold air ahead of the front can subtend the warmer air ahead, and the warmer air being lifted (and hence cooled) will cause condensation of the moisture because colder air cannot hold as much moisture as the warmer air. (I know some may contend that technically that is not correct, but for practical purposes it is.) That is how you get an all-day rain day.