• Hello, please take a minute to check out our awesome content, contributed by the wonderful members of our community. We hope you'll add your own thoughts and opinions by making a free account!

Misc General Banter Thread

Rain Cold said:
And I really don't like probability forecasting. I know some do, but it's basically just taking available model guidance and running it through a math equation to arrive at a number that you can slap in a "forecast".

I don't believe it's a ton more helpful to the public to say there's a 30% probability (vs 40% or 50%) of an inch of snow and a 15% probability (vs 25% or 35%) of 4" of snow on Sunday than it is to say, we're watching the chance of snow on Sunday. Right now, it's most likely that if anything falls, it will be on the low end...around an inch or two. But that could change, so stay tuned.
Click to expand...

I agree I don’t either, it’s just model regurgitation without much additional human input. Garbage in, garbage out.

If these newer AI forecast systems were actually smart enough to properly account for known NWP biases in certain situations, that would be helpful, but we’re clearly not there yet.
 
WolfpackHomer91 said:
He don’t care…. His annual “Breckenridge” Sausage fest he and Ol Coke head Cantore go to each year is coming up. He’s mentally checked out waiting to snuggle and do lines off Jim’s Chest.


Sent from my iPhone using Tapatalk
Click to expand...
There's still time to delete this
 
Webberweather53 said:
I agree I don’t either, it’s just model regurgitation without much additional human input. Garbage in, garbage out.

If these newer AI forecast systems were actually smart enough to properly account for known NWP biases in certain situations, that would be helpful, but we’re clearly not there yet.
Click to expand...
Yeah that's it. A forecast should take input from the models but then go further and draw upon one's expertise and experience and the overall horse sense of the situation. The EPS probabilities that can up or down drastically over the course of 2 model cycles is not a better or more informative method, IMO. Might as well ust load those into a weather app, put your feet up and call it a day.
 
Parker said:
Look, not for those who live there lol. I just associate the beach with hot weather and water. Not a knock on those who live down there and enjoyed a generational event.
Click to expand...
Hey we don’t want it. I moved south for hurricanes not snow and I’ve seen more snow than rain from a tropical system in 3 years.
 
As we should all know by now, a lot can change in 2 days. Usually not for the best but I say we don’t have a good grasp on what’s going to happen with this storm until Sat morning.
 
bouncycorn said:
This is a very interesting discussion to be had!

Probabilistic forecasting is important for meteorologists who are proving context to meteorologists, like NWS briefing TV Mets.. but probabilities are very confusing to the general public. Most, especially the uneducated, just want to know WHAT will happen (deterministic) while the atmosphere is inherently a chaotic, probabilistic system.

How do you communicate literally probabilistic elements to a public that requires deterministic forecasts? It is a tough challenge. Many just go with the probabilities to cover their rear ends, but it may not be the best way to communicate with most of the general public.
Click to expand...
Have you held a position communicating to the general public before? I'm sure it's tough when dealing with those that take forecasts personally.
 
Whos got those good JMA maps? It was victorious 3 days out in dec 2017, then 2days out other models started hinting. But also then the globals for the most part had cold rain for most.
 
You must log in or register to reply here.
Back
Top