December 2025 regional war/obs/disco thread

[CoastalWx]

AI models certainly would. 

[ineedsnow]

10 minutes ago, Damage In Tolland said:

Should be enough to give white Cmas 

Yup we take!! 

[Sey-Mour Snow]

GFS and Euro really beefing up qpf for Tuesday morning .. nice .2-.4” qpf event with some spots up to .5”

[CoastalWx]

The redevelopment on the AI models is nice to see. Would keep flow from ripping out of the SW at 850 and below. Seems like op runs so far are slowly migrating to that idea.

[CoastalWx]

6z euro more with the inv trough look into srn Me and central NH. 

[UnitedWx]

1 hour ago, ineedsnow said:

Yup we take!! 

Do you still have anything on the ground up there? Surprisingly didn't loose everything here... but most, probably 20 percent coverage 

[SouthCoastMA]

57 minutes ago, CoastalWx said:

6z euro more with the inv trough look into srn Me and central NH. 

clips the ema coast Christmas eve morning. It sort of looks like the trough on 12/14, but more pronounced. something to watch at least

[Baroclinic Zone]

LFG!

 

[dendrite]

59 minutes ago, CoastalWx said:

6z euro more with the inv trough look into srn Me and central NH. 

Sweet look for here. GFS works too. Just like the troughs I used to know.

[Typhoon Tip]

2 hours ago, CoastalWx said:

The redevelopment on the AI models is nice to see. Would keep flow from ripping out of the SW at 850 and below. Seems like op runs so far are slowly migrating to that idea.

I'm testing the AI cluster with this 23rd thing ...

Firstly, they're remarkably similar - which I find interesting, actually.   It makes more questions for me.  How is AI GFS and AI Euro, non-dispersive relative to each other?  that's a weirdness.  

I still don't know more precisely how these models are constructed.   What is actually processing.    Until I do, it's voodoo ... Are they just probing the 'cloud' ( ugly pun) for the mean of a billion minds of both machine and man?  If that's the case, of course they are always more encouraging for snow nearing Xmas, because it's tapping into the hopes and dreams of the latter.  Heh.   Or are they processing actual fluid and thermodynamic equations in the propagating of wave mechanics through a matrix.  Something in between?   

I tried researching this question but I'm hit with a typical smoke-screen of marketing and promotional speech - or at least too much and losing patients.. Comes off a bit Wizard Of Ozian and the "don't pay attention to the man behind the curtain".   Typical of the web though.  The entire degenerative irony of the greatest invention of modern industry is that it's usefulness is run down by capital interest and greed.  It's like just answer the fuckin question! no, you're not getting any money you petty f*n  c*ts    

I digress.   Merry Xmas.  LOL    

[cleetussnow]

10 minutes ago, Typhoon Tip said:

I'm testing the AI cluster with this 23rd thing ...

Firstly, they're remarkably similar - which I find interesting, actually.   It makes more questions for me.  How is AI GFS and AI Euro, non-dispersive relative to each other?  that's a weirdness.  

I still don't know more precisely how these models are constructed.   What is actually processing.    Until I do, it's voodoo ... Are they just probing the 'cloud' ( ugly pun) for the mean of a billion minds of both machine and man?  If that's the case, of course they are always more encouraging for snow nearing Xmas, because it's tapping into the hopes and dreams of the latter.  Heh.   Or are they processing actual fluid and thermodynamic equations in the propagating of wave mechanics through a matrix.  Something in between?   

I tried researching this question but I'm hit with a typical smoke-screen of marketing and promotional speech - or at least too much and losing patients.. Comes off a bit Wizard Of Ozian and the "don't pay attention to the man behind the curtain".   Typical of the web though.  The entire degenerative irony of the greatest invention of modern industry is that it's usefulness is run down by capital interest and greed.  It's like just answer the fuckin question! no, you're not getting any money you petty f*n  c*ts    

I digress.   Merry Xmas.  LOL    

Do androids dream of electric sheep?  

[CoastalWx]

18 minutes ago, Typhoon Tip said:

I'm testing the AI cluster with this 23rd thing ...

Firstly, they're remarkably similar - which I find interesting, actually.   It makes more questions for me.  How is AI GFS and AI Euro, non-dispersive relative to each other?  that's a weirdness.  

I still don't know more precisely how these models are constructed.   What is actually processing.    Until I do, it's voodoo ... Are they just probing the 'cloud' ( ugly pun) for the mean of a billion minds of both machine and man?  If that's the case, of course they are always more encouraging for snow nearing Xmas, because it's tapping into the hopes and dreams of the latter.  Heh.   Or are they processing actual fluid and thermodynamic equations in the propagating of wave mechanics through a matrix.  Something in between?   

I tried researching this question but I'm hit with a typical smoke-screen of marketing and promotional speech - or at least too much and losing patients.. Comes off a bit Wizard Of Ozian and the "don't pay attention to the man behind the curtain".   Typical of the web though.  The entire degenerative irony of the greatest invention of modern industry is that it's usefulness is run down by capital interest and greed.  It's like just answer the fuckin question! no, you're not getting any money you petty f*n  c*ts    

I digress.   Merry Xmas.  LOL    

Well now you have me wanting to look again, but as I understand it, there are no physical equations involved like our current computer models. It’s learning from ERA data sets from the past and final outcomes. 

[Damage In Tolland]

Something tells me the Euro is going to fail on Xmas Day 

https://x.com/dhtheweathernut/status/2002384867716108776?s=46&t=dhcbvkjmRcyBVQtDxJ3lRg

[rclab]

18 minutes ago, cleetussnow said:

Do androids dream of electric sheep?  

Only when freed from the surge protector. As always……

[Ginx snewx]

13 minutes ago, CoastalWx said:

Well now you have me wanting to look again, but as I understand it, there are no physical equations involved like our current computer models. It’s learning from ERA data sets from the past and final outcomes. 

Really easy to find.

https://weathersats.com/the-evolution-and-future-of-ai-in-weather-forecasting/#:~:text=AI Models Enhancing Numerical Weather,forecasts more frequently and accurately.

Normalization and Formatting: Data, often in GRIB or NetCDF formats, is cleaned and structured into consistent grids, frequently at 0.25-degree resolution.

Feature Engineering: Raw data is converted into actionable features, such as calculating "feels like" temperature or creating precipitation intensity categories. 

Data Assimilation for "Real-Time" Updates

AI models use advanced techniques to ingest the current state of the atmosphere into the model to begin a forecast: 

AI-Driven Assimilation: Instead of traditional, slow numerical methods, AI accelerates the integration of new observations by filling in data gaps in regions where measurements are sparse.

Hybrid Approaches: Some models, such as NeuralGCM, blend AI with traditional physical simulations to ensure the input data respects atmospheric dynamics.

Specialized Adapters: To shift from historical training data to live data, models may use specialized neural network architectures (like U-Net) to map real-time, "messy" data into the same format used for training. 

Key Technologies Used

Neural Networks (CNNs/RNNs): Convolutional Neural Networks analyze spatial satellite imagery, while Recurrent Neural Networks process time-series data.

Cloud Platforms: Models are often hosted on AWS, Google Cloud, or Microsoft Azure to manage the high volume of data.

Transformers: Used to process long-range dependencies in atmospheric data. 

[RUNNAWAYICEBERG]

LR did not look good overnight. Se ridge connecting with Atlantic ridge while troughs dig into the NW. That stubborn Aleutian ridge is relentless and not helping…

[RUNNAWAYICEBERG]

Pray this fades over time:

[Krs4Lfe]

11 minutes ago, RUNNAWAYICEBERG said:

Pray this fades over time:

This would be the death knells for Winter 2025-2026. Prime time when temps are the coldest, and there is a trough in west Canada and a dip in the jet stream across west/central US, and a ridge over east US. And remember, this is a smoothed ensemble mean. Once we get closer to the date, it won't be as smoothed and the differences between the trough in the west and the ridge in the east would likely be much greater. While the -NAO continues to improve, all other teleconnections remain where we don't want it, including a deepening -PNA. Without help from the PNA and the pacific, the -NAO will be almost meaningless

[mahk_webstah]

6 hours ago, CoastalWx said:

Looks warm around here. Close the shades look until late month.

But if I close the shades, then won’t I be unable to look at the 2 to 4 inches of snow that will fall on Tuesday?

[mahk_webstah]

4 minutes ago, Krs4Lfe said:

This would be the death knells for Winter 2025-2026. Prime time when temps are the coldest, and there is a trough in west Canada and a dip in the jet stream across west/central US, and a ridge over east US. And remember, this is a smoothed ensemble mean. Once we get closer to the date, it won't be as smoothed and the differences between the trough in the west and the ridge in the east would likely be much greater. While the -NAO continues to improve, all other teleconnections remain where we don't want it, including a deepening -PNA. Without help from the PNA and the pacific, the -NAO will be almost meaningless

That might not be too bad up here. I think that’s the pattern we had was it in 2023? We ended up with close to 100 inches here but with a lot of wet, heavy storms where we were right on the edge and you were raining down south

[CoastalWx]

Euro is decent for the 26th NoP.

[Damage In Tolland]

Icy 

[Krs4Lfe]

33 minutes ago, mahk_webstah said:

That might not be too bad up here. I think that’s the pattern we had was it in 2023? We ended up with close to 100 inches here but with a lot of wet, heavy storms where we were right on the edge and you were raining down south

That winter will always live in infamy here in NYC. Cold December, but no snow, and then a torch for rest of the season, with only like 3" of snow in February and a dusting in March. Northern new england through great lakes and out west had a great season though. Lots of record snows across Great Lakes, Upper Midwest, and out west. 

[CoastalWx]

Boxing Day interesting on EPS too

[Fozz]

5 minutes ago, Krs4Lfe said:

That winter will always live in infamy here in NYC. Cold December, but no snow, and then a torch for rest of the season, with only like 3" of snow in February and a dusting in March. Northern new england through great lakes and out west had a great season though. Lots of record snows across Great Lakes, Upper Midwest, and out west. 

I did two ski trips out west that season. It was an all-timer at Alta/Snowbird and the snow was phenomenal.

Needless to say, this better not be a repeat, but I’m skeptical of ensemble means that are 2-3 weeks away.

[Boston Bulldog]

28 minutes ago, CoastalWx said:

Euro is decent for the 26th NoP.

Model whiplash in full effect for the 26th. Euro was plowing this thing across James Bay recently 

[CoastalWx]

Just now, Boston Bulldog said:

Model whiplash in full effect for the 26th. Euro was plowing this thing across James Bay recently 

I know. So much PTSD since 2022, but we hope and pray it has the right idea. 

[WxWatcher007]

3 minutes ago, Boston Bulldog said:

Model whiplash in full effect for the 26th. Euro was plowing this thing across James Bay recently 

My lord we pray it’s right. We’d get smoked here through the 26th.

[Boston Bulldog]

Tip was spot on about the shortwave grapeshot earlier this month wreaking havoc on the models. Different synoptic setup here, but still a relentless barrage of weakish vorticity packets passing through the region. I don’t think we’re near the final solutions for post-Christmas.

That said, the late month signal should coalesce into its final form with more lead time than the medium range events did/will.

[CoastalWx]

5 minutes ago, Boston Bulldog said:

Tip was spot on about the shortwave grapeshot earlier this month wreaking havoc on the models. Different synoptic setup here, but still a relentless barrage of weakish vorticity packets passing through the region. I don’t think we’re near the final solutions for post-Christmas.

That said, the late month signal should coalesce into its final form with more lead time than the medium range events did/will.

A little Hudson Bay ridge tries to form for Boxing Day. Seems like a change, we hope it’s right.