How do they estimate probability of precipitation?

[coriolis]

Lets say that if there are scattered showers moving west to east, and at any given time, the showers cover 50% of the map. Over time the entire area should see some rain, and as I see it, there would be close to a 100% chance that any given location would get rain at some point in time during that period. So they would say 100% POP, right?

Or if there was a narrow band of rain along a front moving west to east. This area of precipitation might only cover 5% of the map, but again, there would be close to a 100% chance that any given location would get rain at some point in time. Again, would that be a 100% POP?

But if the scattered showers moving west to east covered 10% of the map, maybe 50% of the areas would see rain during the period. So in this scenario would they say a 10% chance of precipitation, or as I see it, would they say a 50% chance of rain?

Is there some numerical or statistical model to calculate POP? Do they input coverage and velocity? Or is it simply the subjective opinion of the forecaster?

[Kelly_Thundercloud]

Hey, good question! I'd like to hear the answer to this too...I'm surprised no one has replied yet. Come on, experts! You got two people hungry for an answer lol.

[coriolis]

Sheesh, with all of the degrees, credentials, years of experience, and initials that people have after their names around here, you'd think that some one would know.

Or is the question really that stupid?

[bob rulz]

I am in no way an expert, but mostly from reading local NWS weather discussions I've always thought it was determined based off of the precipitable water content of the atmosphere (usually determined by models, of course). Climatology and various other more minor conditions are taken into account as well (from how I understand it).

And I rarely ever see 100% pop, which I always thought odd since even pretty good-sized storms with nearly complete coverage often come up as 90% pop (at least here). Maybe they're just being cautious? And if coverage is 50% with showers moving through the area, that in no way guarantees rain, although it almost does.

[wall_cloud]

It all depends on how one sees probability of precipitation. First, I would say that a majority of people don't understand it in the first place. I don't know how many times I've heard someone say "The weatherman said it was going to rain" with 20% PoPs and criticize the forecast when an isolated storm did not affect them. Other times, I have forecast 80% + PoPs and only had a few hundredths of an inch of precip and had the forecast criticized. Both forecasts are correct in those instances, but the public doesn't get it.

There are two ways to look at PoPs: Areal vs. Probability. Most forecasters (IMHO) probably view pops as areal in the short term when the decision of whether or not precip will develop has been made. In that scenario, areal qualifiers are used and PoPs become a bit less important. The forecast will read scattered thunderstorms for 30,40,50 percent, likely for 60,70 percent and prevailing for 80 percent and higher. If the radar image roughly fits the qualifier, updates are usually not made. So in that instance, there really isn't that much difference in 30 percent and 50 percent. In the forecast beyond the 2nd period or so, PoPs are more probability based, meaning what is the chance that it rains in any particular location.

Let it be known that NWS precipitation forecasts are based on whether or not it precipitates 0.01" or greater in an official rain gauge. That is what makes forecasters more conservative as a whole. It can rain over the entire forecast area but if it doesn't hit that rain gauge, a dry forecast verifies best. Additionally, if it rains lightly for hours and it never accumulates to 0.01", the observation was considered dry. PoP verification is one of my biggest pet peeves.

I hope that helps.

[coriolis]

Yes that helps, thanks.

My whole concern is whether I'll get rained on if I bike to work. Last Friday I had storm chasing me all the way home, on a 30% POP day. My rule of thumb is that I'll bike it when it's 30% or less.

I guess that if I consistently go out on days with a 30% POP, there's nearly a 100% chance that sooner or later the odds will catch up to me and I'll get wet.

One thing I can do is watch the radar and try to scoot between the storms.

[bob rulz]

It seems around here it usually rains with a 40% pop, and then always rains when there's at least 60%.

[wxman57]

Now you see why I've always hated to quantify precipitation chances with a probability value. Not many people can really define what a 30% chance of rain really means.

I remember taking a course on statistical meteorology back at A&M in the late 1970s. We talked about how to compute the chance of rain for a single point. It's actually the combination of two probabilities. First, there's the question as to whether or not there will be conditions favorable to support any precipitation at all. Then there's the coverage factor to consider.

Generally, a 30% chance of rain means that a point has about a 1 in 3 chance of receiving any precipitation on that day. A higher percent chance of rain doesn't mean greater precipitation amounts are likely, either. Could be a 100% chance of rain and all we're talking about is very light stratoform rain and/or drizzle covering a large area through the day.