March 7th 2018 Coastal Storm thread (not obs)

[strgazr27]

2 minutes ago, Drz1111 said:

This is bad forecasting and frankly beneath you, Don.  Climatology shouldn't enter into a short term forecast like this because it is already accounted for in the models.  Put differently, the model already knows the water temp, the sun angle, the antecedent airmass temperature and still shows the threat; correcting for climatology is double-counting.

Normally when you post, you're talking about long range forecasts.  There, climatology is relevant because all long range model forecasts should be regressed to the mean and climatology is the mean you regress to.  But a short term forecast shouldn't be regressed because the certainty is much, much higher (and to the extent it should be regressed, it needs to be regressed based on the initial conditions - i.e. you wouldn't assume a storm would track west of the model output because it's March, because spring storms don't 'want' to hug the coast unless induced to by the upper air pattern.

Notwithstanding all of this, 4-8" probably an entirely reasonable forecast as a blended mean informed by meteorology, but historic climatology isn't an appropriate or scientific adjustment to the modeled forecast.  

Which is why they’re wrong sometimes. I see nothing wrong with dons post or his call. He’s one of the VERY FEW on here worth listening to. Personally I think 8” for NYC or any place on LI may be wishful thinking. Call it bad forecasting if you want but model runs, trends, local climatology and yes, history all play into my call. FWIW

[Blizzard-on-GFS]

UGLY RGEM run. There’s gonna be a lot of weenie tears in central Jersey if this trend keeps up. Let’s see if the HDRPS agrees. 

[strgazr27]

3 minutes ago, NJwx85 said:

Yeah but that's directly tied to the mid-level centers. If they pass further offshore, both mixing and dry slotting become less of a threat.

Agreed, but at this time I don’t see it shifting that Far East. That’s all I was saying. 

[USCG RS]

The last three runs of the HRRR for 12Z tomorrow have shifted the surface low at least a few miles east at the same latitude each run.  Maybe noise, maybe wipers, maybe a slightly faster track off the coast before the turn.

Imo, nothing is off the table yet.

Sent from my SM-N950U using Tapatalk

[purduewx80]

unshockingly, 18Z GFS trending east/colder.

[weatherpruf]

2 minutes ago, Rtd208 said:

Honestly it looks almost unanimous on the models that the NYC metro will see less snow then what was thought earlier. But once again a highly fluid situation where now casting will be the utmost importance. I still think the NYC metro will see several inches of snow but may not make the double digits. Time will tell. 

 I hope we can salvage a good 4-8. And hopefully it isn't a compacted mess.

[MJO812]

Rgem is great for NYC

[Sn0waddict]

Gfs is definitely better. Not as tucked. Still wish it was a little further south ( its just south of LI) but any improvement is a good one.

[NJwx85]

Looks like the 18z GFS is all rain for NYC, again, I wouldn't give this much weight with regards to thermal profiles.

 

[Drz1111]

5 minutes ago, strgazr27 said:

Which is why they’re wrong sometimes. I see nothing wrong with dons post or his call. He’s one of the VERY FEW on here worth listening to. Personally I think 8” for NYC or any place on LI may be wishful thinking. Call it bad forecasting if you want but model runs, trends, local climatology and yes, history all play into my call. FWIW

"Trends" don't exist.  If the last run of a model moved a storm track east, it doesn't make the next run more likely to go further east.  Models don't 'correct' in that way.  

[Blizzard-on-GFS]

Just now, Drz1111 said:

"Trends" don't exist.  If the last run of a model moved a storm track east, it doesn't make the next run more likely to go east.  Models don't 'correct' in that way.  

People say that but experience proves otherwise. 

[SnowGoose69]

The RGEM really hasn’t moved much the last 4 runs 

[weatherlogix]

8 minutes ago, NJwx85 said:

Looks like the 18z GFS is all rain for NYC, again, I wouldn't give this much weight with regards to thermal profiles.

 

Not really. Here is the KJFK sounding. Look at the bottom right. Best guess on precip (even though the precip maps have RAIN)

[weatherlogix]

Same with 27...even though the precip maps have rain

[NJwx85]

1 minute ago, weatherlogix said:

Not really. Here is the KJFK sounding. Look at the bottom right. Best guess on precip (even though the precip maps have RAIN)

I see what you're saying but temps are a good 2-4 degrees above freezing aren't they? If that verified, it would be a lot of white rain based on that sounding I believe.

[EastonSN+]

Even though it starts out way west, the snowfall maps are basically the same for the GENERAL area as 12z. 

Amazing that this trend is not stopping at all. GFS going to pull the upset?

[weatherpruf]

4 minutes ago, strgazr27 said:

Which is why they’re wrong sometimes. I see nothing wrong with dons post or his call. He’s one of the VERY FEW on here worth listening to. Personally I think 8” for NYC or any place on LI may be wishful thinking. Call it bad forecasting if you want but model runs, trends, local climatology and yes, history all play into my call. FWIW

I can't say and will have to take his word for it, but I have seen a pro met here last year do precisely that  and he nailed the March bust, not ignoring the science ( he was seeing the warm layers ) but also looking at history of March storms. I can't wade into it too much because I don't know the science. Didn't mean to put Don on the spot so much. 

[weatherlogix]

6 minutes ago, NJwx85 said:

I see what you're saying but temps are a good 2-4 degrees above freezing aren't they? If that verified, it would be a lot of white rain based on that sounding I believe.

Ironically and in typical GFS fashion at 30, with the temp 33 and the rest of the column cold it says best guess: RAIN

[MJO812]

2 minutes ago, EastonSN+ said:

Even though it starts out way west, the snowfall maps are basically the same for the GENERAL area as 12z. 

Amazing that this trend is not stopping at all. GFS going to pull the upset?

What ?

It's way better than 12z

[weatherpruf]

8 minutes ago, Drz1111 said:

"Trends" don't exist.  If the last run of a model moved a storm track east, it doesn't make the next run more likely to go further east.  Models don't 'correct' in that way.  

So what are the models useful for? Serious question, not being snide.

[donsutherland1]

8 minutes ago, strgazr27 said:

Which is why they’re wrong sometimes. I see nothing wrong with dons post or his call. He’s one of the VERY FEW on here worth listening to. Personally I think 8” for NYC or any place on LI may be wishful thinking. Call it bad forecasting if you want but model runs, trends, local climatology and yes, history all play into my call. FWIW

@DRZ1111 makes a fair point, IMO.

I use climatology as a reference in weighing uncertainty when there is a large spread in the guidance. It reminds me to go back to look at the maps for various outcomes e.g., the 1960 case, which are available on NCEP's re-analysis page back to the late 19th century, hence my reference about the 1941 and 1960 storm's having colder air masses. I don't literally use climatology in the short-term, but just to go back to look at various outcomes.

[NJwx85]

1 minute ago, weatherlogix said:

Ironically and in typical GFS fashion at 30, with the temp 33 and the rest of the column cold it says best guess: RAIN

I just want to reiterate that I never said I bought the GFS thermals. 

[Zelocita Weather]

GFS looking better, as is the RGEM. Nice slight eastward tweak with the 18z suite so far, but really just noise. 

Also looks like the initial overrunning is more significant than thought. By 7-8am from just pure over running 2-3" could already be on the ground. 

[Drz1111]

7 minutes ago, Blizzard-on-GFS said:

People say that but experience proves otherwise. 

"People" don't say that.   Ask any of the folks on here who professionally develop models.   It's been more than a decade since I was taught this, but my understanding is that folks have tried to prove the existence of trends and have failed.  Your 'experience' is a function of bias - when models random walk in an obvious way, it doesn't stick in your head, but a 'trend' does.  The human mind overfits patterns.

[Stormlover74]

1 minute ago, TriPol said:

UGLY RGEM run/RGEM hasn't changed, GFS is East and Colder/GFS is West and Warmer

How is it possible for people to look at the same picture, yet have completely different interpretations??

Listen, it's March and you're most likely going to get more than 6" unless you live on the South Shore and East End of LI. Be happy.

People live in different places. Rgem cut back on the southern edge 

[NJwx85]

Not sure about precip types yet but 18z RGEM has roughly 45-50 mm of precip on the metrogram for NYC with more to the N & W.

[ILoveWinter]

Honestly the chaos with each model run is very expected, it wouldn’t be a weenie forum without it!

[NJwx85]

3 minutes ago, TriPol said:

UGLY RGEM run/RGEM hasn't changed, GFS is East and Colder/GFS is West and Warmer

How is it possible for people to look at the same picture, yet have completely different interpretations??

Listen, it's March and you're most likely going to get more than 6" unless you live on the South Shore and East End of LI. Be happy.

Depends on location, 18z increased my totals by 3-5”.

[Ericjcrash]

Got the shops closed tomorrow text so it's totally going to rain. GFS ticking easy is encouraging. 

Hi res RGEM less amped thru 17

[tamarack]

8 minutes ago, donsutherland1 said:

@DRZ1111 makes a fair point, IMO.

I use climatology as a reference in weighing uncertainty when there is a large spread in the guidance. It reminds me to go back to look at the maps for various outcomes e.g., the 1960 case, which are available on NCEP's re-analysis page back to the late 19th century, hence my reference about the 1941 and 1960 storm's having colder air masses. I don't literally use climatology in the short-term, but just to go back to look at various outcomes.

Sort of an understatement.  The current rotting little HP means p-type issues all the way into Maine.  March 1960 was the first major snowstorm of my experience where the temps remained under 20° (in NNJ - Jersey Highlands.)  Of course we did it again that December, and even colder for the JFK inaugural storm.  While shoveling the Feb blizzard the air felt almost warm - upper 20s instead of low teens - despite the howling wind.