Potential extensive winter event, I-95 west and with again a chance for NYC first inch(es) of snow Mon or more likely Tue Jan 16, 2024 (serves as OBS thread as well)

[mannynyc]

[MJO812]

Eps shifted west

[MegaMike]

20 minutes ago, wdrag said:

I can only presume AI assisted modeling was tested for improved guidance prior to its implementation.  Recommend staying with WiC and check their afternoon EPD when it comes out

I'm not sure why anyone would use AI to forecast the weather.

1) AI is a glorified analog so it won't perform well for anomalous events.

2) We don't know what's incorporated in AI models so we can't correct for known biases.

3) Why use AI when we already have trusted NWP models?

AI can be used to improve the accuracy of NWP output that have a known, and predictable bias though. For example, improving wind speed: < https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2018JD029590 >.

[vegan_edible]

20 minutes ago, MegaMike said:

I'm not sure why anyone would use AI to forecast the weather.

1) AI is a glorified analog so it won't perform well for anomalous events.

2) We don't know what's incorporated in AI models so we can't correct for known biases.

3) Why use AI when we already have trusted NWP models?

AI can be used to improve the accuracy of NWP output that have a known, and predictable bias though. For example, improving wind speed: < https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2018JD029590 >.

just dropping into say that i'm from latham originally and it warms my heart to see someone on this board from the 518

[North and West]

20 minutes ago, MegaMike said:

I'm not sure why anyone would use AI to forecast the weather.

1) AI is a glorified analog so it won't perform well for anomalous events.

2) We don't know what's incorporated in AI models so we can't correct for known biases.

3) Why use AI when we already have trusted NWP models?

AI can be used to improve the accuracy of NWP output that have a known, and predictable bias though. For example, improving wind speed: < https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2018JD029590 >.

my brother in christ, they're trying to takeover the world

[DenvilleCyclone]

Storm or no storm I took advantage today and took down the Christmas lights.  No way in hell I was going to do it in the cold.  

[kingbaus]

It's almost happy hour time.

[the_other_guy]

As soon as Mike Woods said models are looking south so just a slight chance of snow I warmed up the snow blower.

[MJO812]

[kingbaus]

Icon looks too amped

[MJO812]

9 minutes ago, kingbaus said:

Icon looks too amped

Yep good thing it isn't a good model

[kingbaus]

So far gfs looks flat hopefully improvements will come.

[Dark Star]

2 hours ago, MegaMike said:

I'm not sure why anyone would use AI to forecast the weather.

1) AI is a glorified analog so it won't perform well for anomalous events.

2) We don't know what's incorporated in AI models so we can't correct for known biases.

3) Why use AI when we already have trusted NWP models?

AI can be used to improve the accuracy of NWP output that have a known, and predictable bias though. For example, improving wind speed: < https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2018JD029590 >.

Trusted models?  AI could be a good thing, especially when you can adjust for errors in the program say at day 3 and correct for biases.  Might improve long range patterns?  I would think the AI would correct for known biases, otherwise, why use AI?  I would think no model would handle an anomalous event well?  Or perhaps instead of AI, have human intervention adjusting at say, day 3 in a long range model to correct for a suspected error?

[wishcast_hater]

As soon as Mike Woods said models are looking south so just a slight chance of snow I warmed up the snow blower.

Lol that guy is still on TV? What a joke, he was on vacation more than at work when I used to watch television.


.

[NJwx85]

The only real change on the GFS through 78hrs is that it is slower with the trough.

[NJwx85]

Moving ahead to day four. Looks like the GFS is caving to the Euro.

The Southern vort is too far out ahead of the trough axis since it's slower and your surface low here is way offshore.

[MJO812]

It's much flatter 

[NEG NAO]

2 minutes ago, NJwx85 said:

The only real change on the GFS through 78hrs is that it is slower with the trough.

18Z and 6Z model runs should be taken with a grain of salt they are using stale Radiosonde data

[uofmiami]

 

[NJwx85]

Just now, NEG NAO said:

18Z and 6Z model runs should be taken with a grain of salt they are using stale Radiosonde data

It's caving to the Euro.

[NJwx85]

Waive bye bye

[NJwx85]

3 minutes ago, uofmiami said:

 

As I have been saying all along, it comes down to whether or not the Southern vort is timed well with the incoming trough. If it's too fast like the 12z ECMWF and now 18z GFS shows, the trough will kick it out. If it's too slow it could phase early and you could get a super amplified trough and rain at the coast. Has to be timed perfectly. Still time for this one but there is no denying that this was a major nod towards the Euro.

[NEG NAO]

2 minutes ago, NJwx85 said:

It's caving to the Euro.

and the 12Z EURO with fresh Radiosonde Data Input made a large move towards the GFS - your point being ?

[kingbaus]

Ughhhh awful

[SBUWX23]

The euro has more snow than the GFS now. Models are nodding at nothing other than uncertainty 

[NJwx85]

1 minute ago, SBUWX23 said:

The euro has more snow than the GFS now. Models are nodding at nothing other than uncertainty 

The Euro has been consistently slower with the trough axis dropping down from the upper Plains. Until this timing issue is fully resolved there will be some uncertainty. The hope is that it is still poorly sampled as @Typhoon Tip alluded too earlier. 

[vegan_edible]

if we get cooked at 0z then i'll start worrying. 18z isn't the best anyway, and we know that

[Allsnow]

King euro 

[vegan_edible]

Just now, Allsnow said:

King euro 

you got 18z euro access

[NJwx85]

3 minutes ago, NEG NAO said:

and the 12Z EURO with fresh Radiosonde Data Input made a large move towards the GFS - your point being ?

I don't know how you could call that a large move towards the GFS. It was a move towards the old GFS but still just a glancing blow. Lets see what 00z has to say.