Weather.com Inconsistencies

[TromboneAl]

Weather.com can give very different forecasts on the regular web site versus the mobile web site. I refreshed these multiple times, and yet they were consistently different. 15% precip probability is very different from 70%, and is the difference between a go and a no-go for a bike ride.

 

[Sarah in SC]

I've noted that as well, but never to the extent your example shows. I wonder why?
We watched the weather like maniacs prior to our party, and it was always different between the two.

 

[devans0]

I live in a city that has its motto as "the city of five seasons". Around here, weather can be tough to forecast. I remember a few times last year that it was raining the cliche cats and dogs, and on the other side of the street was sunny and completely dry.
If forecasting was easy, we would ALL be fire'd. I do see how some of the local stations seem to always have exciting long range forecasts, but reduce their forecast to the mean as it gets close to occurring. Sweeps week is the worst.

 

[rmark]

I don't know about weather.com, but some years ago the TV weatherman said they followed three different models, and most days they agreed with each other, but when they didn't is was a best guess as to which one was right.

 

[W2R]

Why not use the National Weather Service page? It gives the % probability of precipitation for each day, anyway, if not for each hour.


For example, here's the NWS page for a location in Northern California. To get the page for the specific location you are interested in, enter it in the box that says "Enter your "City, ST", or zip code".

 

[TromboneAl]

I've always assumed that all the weather reporting sites/TV programs base their reports on the same NWS data. True?

 

[W2R]

I've always assumed that all the weather reporting sites/TV programs base their reports on the same NWS data. True?

No, not at all - - at least that is a huge oversimplification of weather forecasting practices. Some do parrot the NWS reports, and some don't, especially around here. Here in hurricane country the weather is a Big Deal, and I always make a point of listening to some of our local reports that give a more independent perspective. Maybe you have heard of our legendary hurricane forecaster, Nash Roberts, who passed away recently. He also did wonderful day-to-day forecasts when there wasn't an impending hurricane. We also have other independent forecasters here.

These independent forecasters use all data available to them, which may be mainly information from the plethora of meteorological stations and buoys publicly available and funded by the NWS, but their interpretation, rudimentary modeling, and reports are independent and do not just parrot the NWS report. In particular, the % chance of rain (which I believe is your interest here) from an independent forecaster could be a little different. I do not think this is the reason for the differences you observed, though.

The NWS is very expert and hopefully provides information that could help you on your bicycle journeys.

 

[TromboneAl]

Well, they're all saying that we are having rain right now, and this is what the weather's been like all afternoon. I think the weather was more accurate when we lived in the Bay Area.

 

[W2R]

Sorry to hear that it's raining!

Now, all we need is a weather site with a prediction for no rain that will magically cause the rain to stop, somehow.

It's sunny, 82, bright and gorgeous here. Everyone can be envious this week. Eventually the oppressive heat and humidity will swallow us up, but not quite yet. One of our better meteorologists says that we'll need to pull our sweaters back out once again in April before summer is here for good. We shall see.

 

[Ed_The_Gypsy]

At the moment I am living in Baku, Azerbaijan. I have links to three weather sites. They all give different results for any given day. I look out the window for my weather.

 

[Tesaje]

I've always assumed that all the weather reporting sites/TV programs base their reports on the same NWS data. True?

No, anybody can do anything they like with the NWS data and forecasts. They even can take them wholesale and claim it as their own giving zero credit to the NWS. The media can do whatever they like with the data and forecasts as they report it. The only thing they agree to do is to put out the warnings as they are sent by the NWS. But even that is just them agreeing to it and the only leverage is they don't want the liability of not doing it.

The data is mostly sourced directly from the NWS regardless of who is making the forecasts. The NWS is using some 5-8 different models now including some from other countries and they are constantly being tweaked. They don't always agree. Predicting the future always will have some errors. Areas with mountains have more localized effects that are harder for models to predict as well. I always use the NWS weather.gov sites and check the radar.

If you go to the hourly weather graph, you get the forecasts for each hour out to a couple of days. Those are more automated but are approved by human forecasters before they get posted. The NWS makes an effort to not swing widely between forecast areas and shift changes. All that is given to you for a mere $5 a year in taxes per taxpayer. I'm fairly expert as I spent 31 years in the NWS.

 

[TromboneAl]

Something that bugs me is that even though the prediction 10 hours out is often wrong, the TV weather now has seven-day forecasts.

 

[Tesaje]

Something that bugs me is that even though the prediction 10 hours out is often wrong, the TV weather now has seven-day forecasts.

Yeah that bugs me even more. It's more like a 7 day guess. We don't really have that kind of skill but several years ago, they decided to throw it out there. The NWS started it.

 

[Nords]

Something that bugs me is that even though the prediction 10 hours out is often wrong, the TV weather now has seven-day forecasts.

Two different programmers tweaking two different models and dreaming of FIRE...