yes, I know when forecasters say there’s x% chance of rain, they mean based on historical data when conditions are similar it has rained x% of the time. maybe it’s just confirmation bias, but it seems like the predicted rain chance never goes up, always down. like today, by me. When I got up TWC said there was to be a 60% chance of rain with the possibility of severe weather this afternoon into the evening. By 3:30, it was down to a 40% chance. Now, it’s 20%. Do things really change so rapidly their models can’t hold together for even an hour?
It’s partly confirmation bias (I’ve certainly seen the percentage increase), and part of it is the wet bias of forecasts.
The farther out, the less certainty in the forecast. So two days out it might be a 30-40% chance of rain, but at five days out it’s 10-60%. Forecasters are generally going to go with the high side of the estimate to avoid pissing off people, so that gets translated as 60% initially and 40% a few days later, even though the average chance consistently remained at 35%.
To quote Kipling, NWS is “a fraud of monstrous size.” In May of 1944, a team of British meteorologists predicted the weather accurately a week in advance. They had no computer models, just knowledge, intelligence, and very good surface maps. I learned to do this in high school. If you want an accurate three day forecast, you need accurate surface maps. NWS has none, does not even know how to make them properly. What you see on the NOAA site is derived from a computer model, and sometimes wildly inaccurate as far as weather fronts and actual locations of low and high pressure areas. The data itself is accurate, but even the isobars do not fit the data, which is itself too sparse to allow decent accuracy of interpretation. I watch the weather (my own weather station) constantly at my location, and monitor weather predictions. They are accurate more than six hours in advance about 40% of the time.
Computer modeling is great, and should be funded and pursued. However, they are trying to model a complex system of a size few, if any, people can even begin to comprehend. Networking every computer on the planet would not be enough to start. Someday it will be, but not for a long time.
I’m sorry, this simply isn’t true. I think you need to define what you mean by not accurate, because 1-day forecasts are much more accurate than 40% of the time. Here’s a better reference than one person reporting based on their home weather station.
And the briefest review of the NWS website would reveal they do indeed have surface maps. Many surface maps, in fact. There’s even a page titled Current Surface Maps.
now I want to know what in my OP would have brought on a new poster with an apparent axe to grind.
It is profoundly annoying. Two days back I looked at the Met site in Britain — apparent nearest site, but that’s a fraud since the county just goes to a few weather-stations — and it promised rain quite heavily for a few hours in the week ahead; just looked again and the predictions are diminished. We rarely get rain here and today’s was an awful 78 F. so it would have been a relief.
It’s like false advertising.
Also, I overlooked the bit in your OP about it changing significantly within a few hours. I’d guess that isn’t typical, but it really depends on your location and the specific weather event. Precipitation in particular is harder to predict. Under certain conditions, the models can change that quickly. But it’s still most likely a change in certainty as the error bars are brought in.
yeah, I’m going to walk this back and go with confirmation bias. 'cos it wasn’t too long after my OP when the rain chance went up to 60% and it started pouring. and it’s still raining.
Wait…what? That don’t make no sense. 78 in the summer would be downright orgasmic.
Seriously though; what I’ve noticed is that generally the entire-day forecasts don’t fluctuate all that drastically, but the hourly ones do, as they re-run the forecasts and get even more up-to-date data.
One thing I got out of my sole meteorology course in college is that the closer the time horizon, the more accurate the forecast. At the time, anything past 3 days was essentially up there with voodoo or divination as far as accuracy was concerned. Inside of 3 days, and they had a decent chance of being right, with the 24 hour forecasts being like 90% correct.
I’m sure with modeling and computing advances, those numbers are higher/further out now, some 25 years later.
I suspect they intentionally use the highest of a range of likely values, in order to not be accused of failing to forecast rain. If best available forecasting tools present a model in which the probability is 20-40%, they will say 40%, because it is better to forecast rain and not have any, then to forecast no rain and have people pissed off that they got wet without expecting to.
Wow, I just stumbled onto this board, and didn’t realize I signed up 5 years ago. Where did all the time go? 
My life-long (emphasize “long”) observations about TV meteorologists:
They have their own culture, and apparently are confined pretty much inside of it. Weather is their job and it seems to become their life.
Take the weather, and the forecasters, here in the northeast. The most boring thing they can forecast in July is “Expect a high of 84 with hazy skies, and a slight chance of an evening thunderstorm.”
Or, equally as boring in January: “We will have a high tomorrow in the upper 20s with a chance of flurries, accumulations of an inch or less.”
ZZZZZZZ, this stuff puts them to sleep. What they WANT to forecast in the summer is "A high in the upper 90s, possibly breaking the record . . . " Or in January, “Expect heavy snow starting tomorrow, with accumulations of 18-24 inches, with wind chills below zero.” These kinds of forecasts animates them, makes them smiley and jumpy all over the TV screen. I see it all the time.
They live for extremes. Average bores them. So they typically overshoot predictions; they seem to be unwittingly biased that way. Temperature extremes – winter or summer – often come out to be less than what they predict.
Another thing about their culture: They must learn in meteorological school to say "And for your Monday . . . " Huh? YOUR Monday? Who talks like this? Nobody I know or ever knew outside of our local TV meteorologists.
Is the weather for “my Monday” going to be different than my neighbor’s?
Your Monday is the Monday that you will experience where you are. If you live in Cape Town, YOUR Christmas will likely be spent at the beach, but not mine. What’s wrong with that usage? “Your”, if I recall my school grammar, is both the singular and the plural second person, although a Texas weather girl might say “y’all’s Monday”.
They’re playing to a local audience, so MY Monday will be the same as anyone else who’s watching.
It’s a meaningless fluff word, heard only within the insular meteorological culture. And it’s relatively new. Nobody said it before 20 or so years ago. It follows the degradation of government schools and what they do (or don’t) teach.
How’s your day been? Bad?
In the 70s and sunny! Great! Screw the weathermen (and bitches). :eek:
Hurk, not entirely sure why you felt the need to call female weather forecasters “bitches”, but how about you dial it back.
Thanks.