Jump to content
  • Member Statistics

    17,502
    Total Members
    7,904
    Most Online
    Weathernoob335
    Newest Member
    Weathernoob335
    Joined

January 20-22 “bring the mojo” winter storm threat


lilj4425
 Share

Recommended Posts

Just now, burgertime said:

Worst part is, it's such a good setup. At least the pattern seems to be reloading so there's always the next one of it his doesn't turn out good. 

Yeah that could be the last shot too. After that we start battling sun angle and so on.

Link to comment
Share on other sites

Just now, JoshWeather said:

i've seen you post this numerous times now.  You've been here for over a decade.  If you think not trusting the models outside of 24 hours(heck, even closer than that) is something new...then you haven't been paying attention.  I couldn't count the number of storms on all my fingers and toes over the past decade that showed a hefty snow only to completely dissipate on the models over the last few days.  

But it wasn't the last few days. It went off the cliff in less than 12 hours. 

Link to comment
Share on other sites

3 minutes ago, burgertime said:

I'm really curious what happens Friday it looks like this could be the rare case where it isn't just cold air chasing moisture especially for CNC and ENC. It may be a little surprise thump.

Also due to the cold front one of the NAM doesn't really mean anything at the end of the day, just looks about right. Still think that shield of precip is probably bigger than NAM is showing but that may not help too much. 

While this sucks for a lot of people this is exactly what we down here in eastern NC like to see in this range.....better cold and not phased and hope for a late bloomer, than watching it amp up and cut and give us rain again....its how we snow big, a lot can still change between now and Friday, and the NAM is all alone right now with this big of a whiff.

 

  • Like 4
Link to comment
Share on other sites

Just now, Brick Tamland said:

Makes me wonder if the globals just don't know how to handle our climate anymore. 

It amazes me that NCEP hasn't incorporated AI into their modeling while virtually every other industry is incorporating it. I'm pretty sure Deep Learning modeling can drastically increase accuracy. 

  • Like 1
Link to comment
Share on other sites

Just now, snowmaker13 said:

It amazes me that NCEP hasn't incorporated AI into their modeling while virtually every other industry is incorporating it. I'm pretty sure Deep Learning modeling can drastically increase accuracy. 

Not to get too far off here but how would one really implement it? Seems like it would be very tough and extremely costly. I'm sure these huge super computers of the globals already use some sort of AI or advance learning. We're dealing with some insane computations. It's probably something that takes years to really implement with such focused AI. 

Link to comment
Share on other sites

Dang I feel bad for the triangle folk. This may not matter but I kept notes on all the model runs for last weeks storm and the NAM was too dry in my backyard in the 72-36hr time frame. The driest model by far on average and it was showing .4-.6 of liquid here a lot of runs and I wound up with 1.2. Fwiw 

  • Like 3
Link to comment
Share on other sites

3 minutes ago, downeastnc said:

While this sucks for a lot of people this is exactly what we down here in eastern NC like to see in this range.....better cold and not phased and hope for a late bloomer, than watching it amp up and cut and give us rain again....its how we snow big, a lot can still change between now and Friday, and the NAM is all alone right now with this big of a whiff.

 

Yea still looks like it could be a fun one for you guys. Also this is just one run.... 

  • Like 1
Link to comment
Share on other sites

8 minutes ago, snowmaker13 said:

Corpus Christi gets slammed this run while we get nothing ... demoralizing. 

namconus_asnow_eus_18.png

They won't see that in the form of Snow, this almost always checks out as sleet or ZR. Plus their ground temps are in the 50s. As a home grown south Texan I can tell you that will be a few slick bridges and tiny ice cycles hanging from street signs. Don't get too affected. They are probably wishing for what NC has a chance at, even if it is small.

  • Like 1
Link to comment
Share on other sites

Just now, burrel2 said:

Dang I feel bad for the triangle folk. This may not matter but I kept notes on all the model runs for last weeks storm and the NAM was too dry in my backyard in the 72-36hr time frame. The driest model by far on average and it was showing .4-.6 of liquid here a lot of runs and I wound up with 1.2. Fwiw 

Also near the end it starts blowing that low up. This is a plausible solution but it doesn't mean it is THE solution. Start the cliff diving if all the models converge around 12z Thursday. 

  • Like 1
Link to comment
Share on other sites

1 minute ago, burgertime said:

Not to get too far off here but how would one really implement it? Seems like it would be very tough and extremely costly. I'm sure these huge super computers of the globals already use some sort of AI or advance learning. We're dealing with some insane computations. It's probably something that takes years to really implement with such focused AI. 

Quantum computing  and super computers has hit major breakthroughs. Even with today's computing technology, you could develop a algorithm. It comes down to staffing. Not many people are AI coders. Hopefully that changes more. 

Link to comment
Share on other sites

1 minute ago, burgertime said:

Also near the end it starts blowing that low up. This is a plausible solution but it doesn't mean it is THE solution. Start the cliff diving if all the models converge around 12z Thursday. 

12z Thursday for a resolution on 24hrs out is a bit insulting from a weather forecast model perspective isn't it?

  • Like 1
Link to comment
Share on other sites

Just now, snowmaker13 said:

Quantum computing  and super computers has hit major breakthroughs. Even with today's computing technology, you could develop a algorithm. It comes down to staffing. Not many people are AI coders. Hopefully that changes more. 

Not sure what field you're in but it's really not that easy. I had to help with a simple AI project with a LOT of budget. First you have to find the people which is difficult. Then there is a lot of trail and error. Simple things like finding discrepancies in finance programs can take a year to implement. Just not sure we're there yet for weather models... not to mention quantum computing is still very much in it's early years. But could be wrong and maybe you have more experience.  

Link to comment
Share on other sites

2 minutes ago, Brick Tamland said:

I mean what's the point in ever looking at the models past 24 hours then?

If you would bother to learn anything you'd know the models are a guide not a map. For 10 years of being on the board you should be able to look at the models, remember how storms have played out in the past and then decide for yourself what is likely to happen. Not live and die by each model run FFS. 

  • Like 10
  • Thanks 6
Link to comment
Share on other sites

9 minutes ago, burgertime said:

Not to get too far off here but how would one really implement it? Seems like it would be very tough and extremely costly. I'm sure these huge super computers of the globals already use some sort of AI or advance learning. We're dealing with some insane computations. It's probably something that takes years to really implement with such focused AI. 

Machine Learning engineers are plentiful. NOAA just doesn't pay well enough - Google and Meta pay the above-average ones $500k+ a year to figure out how to get your meemaw to click more ads.

Link to comment
Share on other sites

4 minutes ago, burgertime said:

If you would bother to learn anything you'd know the models are a guide not a map. For 10 years of being on the board you should be able to look at the models, remember how storms have played out in the past and then decide for yourself what is likely to happen. Not live and die by each model run FFS. 

It's not just one model run, though. The storm has continued to look worse and worse today with each run after looking awesome just 12 hours ago, and only 48 hours before go time. It's the way the runs have gone the opposite direction so quickly and so close to go time that is crazy. Sorry, but that is an epic collapse by the models. 

  • Thanks 1
  • Weenie 3
Link to comment
Share on other sites

2 minutes ago, burgertime said:

Not sure what field you're in but it's really not that easy. I had to help with a simple AI project with a LOT of budget. First you have to find the people which is difficult. Then there is a lot of trail and error. Simple things like finding discrepancies in finance programs can take a year to implement. Just not sure we're there yet for weather models... not to mention quantum computing is still very much in it's early years. But could be wrong and maybe you have more experience.  

I do some AI in my field of work and wrote algorithms. It's not that memory intensive, especially with NCEP's existing infrastructure. Deep Learning can help. 

Link to comment
Share on other sites

1 minute ago, a5ehren said:

Machine Learning engineers are plentiful. NOAA just doesn't pay well enough - Google and Meta pay the above-average ones $500k+ a year to figure out how to get your meemaw to click more ads.

Its hard, or should I say impossible, for the government to compete against the private sector.

Link to comment
Share on other sites

22 minutes ago, Brick Tamland said:

From a huge storm on the global models to nothing here in less than 12 hours and only 48 hours out. Going to be hard to trust the models at all outside 24 hours after this one. This is a Falcons against the Pats in the Super Bowl level of collapse. 

Brick I love your enthusiasm bro but come on stop posting this man lol

  • Like 8
Link to comment
Share on other sites

2 minutes ago, burgertime said:

Not sure what field you're in but it's really not that easy. I had to help with a simple AI project with a LOT of budget. First you have to find the people which is difficult. Then there is a lot of trail and error. Simple things like finding discrepancies in finance programs can take a year to implement. Just not sure we're there yet for weather models... not to mention quantum computing is still very much in it's early years. But could be wrong and maybe you have more experience.  

Yes and just should keep things on perspective as to just how much we understand about our Atmosphere, not much. There are so many variable parameters that have to be collated into a cohesive solution and that does not even include the myraid of things we don't even know yet about our stochastic atmosphere

  • Like 3
Link to comment
Share on other sites

1 minute ago, a5ehren said:

Machine Learning engineers are plentiful. NOAA just doesn't pay well enough - Google and Meta pay the above-average ones $500k+ a year to figure out how to get your meemaw to click more ads.

+500k ...how can NOAA even attempt to match that? They only got funds for the new GFS super computers due to super storm Sandy.  Also machine learning on ads is vastly different than models that use ridiculously sophisticated math and massive data input (though it may be similar to the amount of data from ads?).  Again you may be 100% right just when I was on a project to do AI it wasn't that simple. By the way there is an interesting article here which may back you up. 

https://venturebeat.com/2021/09/29/deepmind-claims-its-ai-weather-forecasting-model-beats-conventional-models/#:~:text=DeepMind claims its AI weather forecasting model beats conventional models,-Kyle Wiggers%40Kyle_L_Wiggers&text=In a paper published in,usefulness in 88% of cases.

  • Like 1
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...