duke_of_hazard

when the weather report says there is a 30% chance of rain, what does that mean? Does it mean there's only 30% chance of water hitting a certain spot over 12 hours? I assume there's some universal standard by which these rain forecasts are made?

For example I never knew that the temperatures are always given in the shade!

Machka 

You can actually look up the answer to that question on your local weather information site:
https://www.weatheroffice.gc.ca/mainm....html#weather8

Q.What is the Probability of Precipitation, or POP?
A.The probability of precipitation (POP) is the chance that measurable precipitation (0.2 mm of rain or 0.2 cm of snow) will fall on any point of the forecast region during the forecast period. For example, a 30% probability of precipitation means that the chance of you getting rained over (or snowed over in winter) is 3 in 10. In other words, there is a 30% chance that rain or snow will fall on you, and, therefore, a 70% chance that it won't. It must also be noted that a low POP does not mean a sunny day: it only means a day where the chance of rain or snow is low.

This fact sheet explains the Probability of Precipitation: https://www.msc.ec.gc.ca/cd/probability_e.cfm.

aRoudy1

A 50% chance of rain means that if you go for a ride, it will rain; if you don't go for a ride, it won't rain.

JimF22003

But here's the metaphysical question. If there's 'x' percent chance of getting rained on if you stay stationary, does it increase or decrease your chances if you're in motion (i.e. on a bike.) Are you moving from one area where you missed the rain to another area where the rain just missed you?

My personal theory is that on a ride, the rain is chasing me like a wheelsucker.

wahoonc

In the USA you can use the NOAA hourly weather graph to get a feel for when the rain will occur and if you read the synopsis you can usually get a good idea of what is triggering the weather, local knowledge helps too.

Aaron

xenologer

Ignore the numbers, Just ride anyway.

duke_of_hazard

The confusing part of this is how to interpret hourly vs. daily forecast . If the daily forecast says 30% chance, I have seen hourly forecast say 30% also for each hour. So over 3 hours wouldn't the probability be 30 * 3 or 90% ?

Ka_Jun

What other job can you have and keep and be only correct 30% of the time?

Condorita

Baseball: a 70% failure rate is considered successful.

wahoonc

Actually they are better than that in many locales...FWIW where I am right now the forecasts are correct 54.7% of the time over a years worth. In SC it was over 80% correct. Forecast Advisor will tell you who gets it right most often.

Aaron

Ka_Jun

must be nice.

127.0.0.1

study here

https://www.arl.noaa.gov/READY_animations.php

SaiKaiTai

From what I recall from my college meteorology course from many years ago it means that, for a given set of same or similar conditions, historically it has rained x% of the time.

I don't know how far back "historically" goes but, essentially, it says -by way of example- "we have seen these conditions 100 times over the past 10 years and it has rained 10 times yielding a 10% chance of rain"

GreenGrasshoppr

If the weatherperson on tv is holding an umbrella and rain is falling, you can trust that person that it is indeed raining.

Nermal

30% is .3, not 3. Why don't you say that .3 X .3 = .09, and it just ain't going to rain?

duke_of_hazard

Lots of replies, but I am not sure anyone has answered the basic question. Over what period of time is the rain chance applicable? I.e isn't 30% of rain over 12 hours mean something completely different than 30% chance of rain in the next hour?

n00bL35

1. Look at charts
2. ???
3. Profit!

chs4

For it to be meaningful it has to be based on whatever period of time the forecast is referring to.

If they say there's a 30% chance of rain on Saturday, I'd take that to mean from 12:00am to 11:59pm on Saturday. Most TV forecasts will break it out into roughly 12hr periods. Overnight and tomorrow, for instance, would be two separate forecasts each with their own assigned probability.

If you go to www.weather.com they'll even give you an hour-by-hour probability of precipitation.

Here in New England I think forecasts are good for about ten minutes, 'cause that's about how long it takes for the weather to change.

Andy_K 

Really, you can hardly ever say the forecast was wrong. If they say there's a 95% chance of rain and it doesn't rain, they still weren't wrong. That's the beauty of probability.

This winter, anything under 50% chance of rain usually meant a dry commute. This spring, anything over 15% has meant I'm getting wet. It's a mystery to me why that happened.

stapfam

I don't need any forecast- If the cloud is sitting over the hills to the South of me- Then it is going to rain shortly. And if I can't see the hills- then it is already raining?

If the forecasters are using previous data to forecast the liklihood of rain- then we have a problem. Weathar patterns have changed so much in the last 5 years that there will not be enough data to give a forecast. The way they work it over here is by satellite forecasts to predict where the rain that is falling in the west- is going to be falling in a couple of hours time. That is fine till they forecast that the rain is going to hit the South East within a set period of time. "The heavy rain that is in the West of the UK will progressively move East to give heavy rain in Sussex by mid morning" was the last one. Except they forget about the change in wind direction and The North of England has Floods because it was not forecast to go there.

deraltekluge

If there's 30% chance of rain in each hour, that's a probability of 0.3 that there'll be measurable rain during that hour. You cannot get the probability of rain in N hours by either adding or multiplying that 0.3

You have to deal with the probability of NOT having rain...

If the probability of rain is 0.3, the probability of no rain is 1 minus that, or 0.7

Then, you can multiply the probabilities of not having rain...for two hours, that'd be 0.7², or 0.49

For three hours, it'd be 0.7³, or 0.343

You can then find the probability of rain by subtracting the probability of no rain from 1.

So, if the probability of rain in a one hour period is 0.3 (30%), the probability in a two hour period is 0.51 (51%), and in a three hour period it's 0.657 (about 66%)

varminter

it means take your rain jacket with you

duke_of_hazard

Great, that was the answer I was looking for!

bmorey

Actually anything other than 100% or 0% chance means 'we took a best guess'. A % chance of rain is essentially useless. The method of forescasting we use in Melbourne ("Three seasons in one day") is to walk outside. If it's raining there's a good chance it will rain. Works everywhere. Weather forecasting is a money-eating boondoggle. Ever never noticed how the 7 day forecast rarely becomes the 6-day, 5-day, &c?