January 2024 Weather in the PNW (Part I)

[MikeInEverett]

This is a dream CZ location. Let’s put it here again 5-6 days from now.

[SnowySeeker50]

[the_convergence_zone]

36 minutes ago, mjreich said:

Not quite in the same way, since they are by definition probabilistic.  The AI models are essentially taking a complex set of atmospheric conditions and turning them into mathematical vectors that can be used to query for similarity - the process of training is itself probabilistic, so you only need run the model once to determine the highest probability outcome for a given set of specific criteria. 

This is one of the fundamental advances in the field. Ensembles only exist to deal with the inherent stochastic limitations of a deterministic physics model; they are trying to capture the probability distribution, but in a very backwards way. We never known whether the ensemble distribution is representative of reality, or just a distribution of the limitations of the underlying physics packages. 
 

AI model solutions generally reflect the probabilities of actual weather conditions occurring, which is why they are/will be so much better than our current model outputs.  Add in the simplicity of adding additional data (adding additional vectors without having to bother with physics whackamole), the infinite geographic scaling and you end up with a completely different category of weather prediction that what we’re familiar with. 
 

Part of the existential crisis in weather forecasting behind closed doors right now is that this technology renders the NBM and much of the existing forecasting infrastructure obsolete, along with a lot of jobs devoted to developing and servicing the large global deterministic models. Lots of competing pressures and entrenched interests. 

This is really interesting. Do you think NCEP is going to pivot at some point or are they going to continue on this path developing FV3-based everything? At least it seems like data assimilation might need more attention? 
 

I work on the observation side and it seems like there’s going to be more demand for our products than ever before, everyone wants to use weather/climate data and use it smarter than they have in the past and automate as much of the decision making as they can. So the overall need for atmospheric science expertise is increasing, people are just going to have to adapt and pivot. 

[Randyc321]

1 hour ago, snow_wizard said:

Awesome run from Portland north.

I like that kind of padding.  (Sorry Portlandians and Portlanders, whichever you identify with) 

[CliffMassYelledAtMe]

39 minutes ago, mjreich said:

Not quite in the same way, since they are by definition probabilistic.  The AI models are essentially taking a complex set of atmospheric conditions and turning them into mathematical vectors that can be used to query for similarity - the process of training is itself probabilistic, so you only need run the model once to determine the highest probability outcome for a given set of specific criteria. 

This is one of the fundamental advances in the field. Ensembles only exist to deal with the inherent stochastic limitations of a deterministic physics model; they are trying to capture the probability distribution, but in a very backwards way. We never known whether the ensemble distribution is representative of reality, or just a distribution of the limitations of the underlying physics packages. 
 

AI model solutions generally reflect the probabilities of actual weather conditions occurring, which is why they are/will be so much better than our current model outputs.  Add in the simplicity of adding additional data (adding additional vectors without having to bother with physics whackamole), the infinite geographic scaling and you end up with a completely different category of weather prediction that what we’re familiar with. 
 

Part of the existential crisis in weather forecasting behind closed doors right now is that this technology renders the NBM and much of the existing forecasting infrastructure obsolete, along with a lot of jobs devoted to developing and servicing the large global deterministic models. Lots of competing pressures and entrenched interests. 

Really stupid question here, but is AI used on the different members of the models? As in using AI to evaluate the accuracy of each member to determine why one member was more accurate than another? And then using that as another source of data in pattern recognition and retraining? 

 

Basically, even with improved AI, is there a purpose for continuing models? 

[TacomaWx]

12 minutes ago, SnarkyGoblin said:

Yo how many years does it take to do that?  They've been expanding that road since I was in college, which is over 20 years ago 

They say that the permanent bridge and the area from I-90 to the top of tiger mountain will be done in 2026. Obviously it’s going to take a lot longer than that. Still so much work to be done. The area on the other side of tiger mountain down to maple valley is a whole second half of the project. 

[Phishy Wx]

thread has digressed to road construction and kids

[Meatyorologist]

When does the 18z Euro release?

[SpaceRace22]

The hurricane hunters took off at 10:45am this morning, so their data will make it into the 00z runs. Expect major changes to tonight's 00zs. 

Considering that we have the Euro, Spire, GEM, UKMET, and GraphCast all firmly on our side, and only the GFS not on our side I think it's clear which model is going to blink here.

https://www.flightaware.com/live/flight/TEAL83

[Doinko]

1 hour ago, TigerWoodsLibido said:

Gonna lowkey be depressed as **** seeing everybody's snow pictures when I haven't had the opportunity to post any daytime snow pictures since 2-25-19. And still zero Arctic airmasses at the upper level in over 7 years...

 

It's just the 18z GFS. The 12z Euro was amazing for your location. The 18z was a step in the right direction as well

 

 

[Hyperbolic Trendz]

3 hours ago, TigerWoodsLibido said:

Less than an hour til a soul-crushing GFS run.

South valley is fuckked.

Always surprised what a defeatist you are considering that epic 24-hour snowfall you had  less than five years ago.  I would have killed for that here in S.Metro where ice and warm-nosing have reigned supreme.

[Cold Snap]

4 minutes ago, Meatyorologist said:

When does the 18z Euro release?

It runs from 3:45-4:15.

[HuskyMaestro]

Throwback to December 25, 2017 of little ol’ Frosty Jr May he bless the models with pepto for the whole region 

[Randyc321]

1 hour ago, TigerWoodsLibido said:

Gonna lowkey be depressed as **** seeing everybody's snow pictures when I haven't had the opportunity to post any daytime snow pictures since 2-25-19. And still zero Arctic airmasses at the upper level in over 7 years...

 

Dude, what do you expect the forum to do?  You get major patience and understanding on here but tone down the woe is me sometimes, OK?  Sometimes you have to buck and not try to bring everyone else down.

Especially before we know what the F is going to happen anyway!

Patience and gratitude for what you got.

Just my thoughts.

[BLI snowman]

Big hail shower just moved through here. Dropped us to  38.

[Slushy Inch]

Don’t judge. I couldn’t resist.

 

[the_convergence_zone]

I was busy today but I took a look at the 12Z…if you ignore everything except the Euro you would probably think that the forecast hasn’t really changed much, just a bit of moderation and maybe 4-5 fewer ensemble members on board. And that may be the reality since the EPS is so far ahead of everything else. 

[Meatyorologist]

PSCZ bearing down on NE Seattle!

[Slushy Inch]

Check out this improvement. I have no idea if it was posted or not. Similar for Boeing field.

[Frontal Snowsquall]

15 minutes ago, SpaceRace22 said:

The hurricane hunters took off at 10:45am this morning, so their data will make it into the 00z runs. Expect major changes to tonight's 00zs. 

Considering that we have the Euro, Spire, GEM, UKMET, and GraphCast all firmly on our side, and only the GFS not on our side I think it's clear which model is going to blink here.

https://www.flightaware.com/live/flight/TEAL83

Boom, let’s go!!!

[Randyc321]

10 minutes ago, BLI snowman said:

Big hail shower just moved through here. Dropped us to  38.

Yep, it's chilly upstairs right now.

Current 850's

[Corncob]

7 minutes ago, Slushy Inch said:

Check out this improvement. I have no idea if it was posted or not. Similar for Boeing field.

Could we see it for PAE Paine Field?

[Meatyorologist]

What a beast!!!

[Randyc321]

2 minutes ago, Meatyorologist said:

What a beast!!!

 

[Phishy Wx]

Thundersnow on Tuesday?

 

 

[MWG]

High of 43F and currently 39F. 

[Phil]

2 hours ago, iFred said:

...and for the ICON.

Three big reasons why it is still worth something in its own right.

It isn't gridded, so it can pick up polar features a bit better than it's counterparts. It isn't hydrostatic, so the "brown water" effect and sea ice are better reflected than the GFS and low res Euro. It also as probably the best global terrain resolution for a "Non-AI" model.

It also has one of the better initialization schemes out there.

[snow_wizard]

Pretty brisk out there today.  Only made 42 for a high.

[Phil]

Lmao, the first tornado of 2024 passed a few hundred yards from where I was staying in FL last week. Trollolol.

[Cascadia_Wx]

Just had a decent graupel shower here. High of just 46 today, and that was at midnight. Afternoon high of 44.

[Deweydog]

Chunky rain!

[Slushy Inch]

Special weather statement for wester WA. Are specials statements coming back? I always loved those. And the nowcasts 

[Slushy Inch]

18z euro is running!

[TT-SEA]

Wow!

[SnarkyGoblin]

3 minutes ago, Slushy Inch said:

Special weather statement for wester WA. Are specials statements coming back? I always loved those. And the nowcasts 

No mention of snow in it.

[RentonHill]

37 minutes ago, Meatyorologist said:

When does the 18z Euro release?

Seriously? You do not deserve the moderator badge. 

[Jbolin]

10 minutes ago, Phil said:

Lmao, the first tornado of 2024 passed a few hundred yards from where I was staying in FL last week. Trollolol.

 

[Cloud]

[Meatyorologist]

Graupel!

[Rubus Leucodermis]

2 minutes ago, Meatyorologist said:

Graupel!

Michael Snyder will be with you shortly.