NYC/PHL Dec 26-27 Potential - Part 6

[weathercoins]

I can't, I'm posting from my ipod right now. That would be pretty tough.

+1

[LVblizzard]

Even at 12 hours on the Euro you can tell it'll probably follow all the other models.

[tombo82685]

hr 12 trof slightly more amplified..h5 opened up over louisiana....continued higher hgts along east coast...sub 1016 low just south of the boot of louisiana

[tombo82685]

hr 18 way a good bit more amplified with the trof...sub 1012 low about 50 miles south of mobile, al...continued hgr hgts along east coast...

[F5TornadoF5]

That's 4am Monday Morning - Long Island is getting pwneddddd.. Do you have the next map? Does it snow till Noon Monday w/ backlash snow? How fast does this storm exit?? Earthlight??

Model is still coming out in 3hr increments. BTW that map is 57hrs.

[masomenos]

The bufkit numbers out of the Hudson Valley are pretty incredible. I know the plight of the HV in these kind of storms all too well and it appears we have a favorable setup to really cash in for once.

[uofmiami]

That's 4am Monday Morning - Long Island is getting pwneddddd.. Do you have the next map? Does it snow till Noon Monday w/ backlash snow? How fast does this storm exit?? Earthlight??

Jeff,

You can check out Suny MM5 yourself here:

http://cheget.msrc.sunysb.edu/html/alt_mm5.cgi

[tombo82685]

hr 24 big time more amplified than 12z....sub 1008 low over the knob of the fl panhandle....lgt precip knocking on dc's door...

[Allsnow]

at 24 hrs so many improvments from 12z...i never seen the euro do this inside 72...wow trough is going negative...

[earthlight]

Already way west of it's 12z run

[Jefflaw77]

thank you

Jeff,

You can check out Suny MM5 yourself here:

http://cheget.msrc.s...tml/alt_mm5.cgi

[RobbTC]

Glenn's forecast just posted

http://www.nbcphiladelphia.com/news/breaking/Changes-in-Snow-Forecast-Sunday-112446074.html

[Allsnow]

hr 30 snowing from atl-dca

[tombo82685]

hr 30 sub 1004 low about 50 miles east of savannah ga....12z had a sub 1004 low about 150-200 miles east of myrtle beach at this time...lgt precip just past dc...lgt to od precip in central va

[RU848789]

At what point do we stop looking at these models and go to purenowcastimg? I know NAM is considered the better "now" model by many, but when do the GFS, euro et al become pointless die to timing?

Never, really - in fact, as one gets closer to the event the models get better and better, as they have less time for the various errors in assumptions/initial conditions/boundary conditions/data input etc. to propagate out until the actual storm event. Think of it this way: almost every model should give nearly 100% accuracy for forecasted conditions 5 minutes into the future, since the future state is so similar to the current state. Put mathematically, the set of non-linear partial differential equations governing the various physicochemical processes occurring in the atmosphere only has to move forward a small time increment, making the range of possible solutions [actually, numerical solutions are the only way to solve these sets of equations, since analytical solutions are simply impossible] much smaller than for a forecast 24 hours into the future.

This major decrease in the forecast error bars (think of a hurricane's trajectory cone, which gets wider and wider as one moves further into the future) as one gets closer to the event is the reason why people have to put a lot more stock in tonight's 00Z runs than they were able to in the "blizzard" runs from 2 days ago. If the Euro is on board and HPC doesn't come up with any reason to throw out tonight's model runs, then I would think we're clearly ready to lock and load on this event (i.e., significant snows are nearly a sure thing for the DC-Boston corridor; blizzard/snowmageddon conditions are not yet a sure thing) and it would then take a nearly unprecedented total model meltdown scenario, a la March 2001, for this to be a complete miss. Of course, I'm not a meteorologist, so I could be off, but I do know a thing or two about numerical models, physical chemistry, and mass, heat and momentum transfer...

[yankeex777]

This is like the old days when you were a kid and just woke up to snow on the ground. I love surprise storms and this is probably the closest we will get to one unless you live without tv or internet.

Yea the Internet has pretty much spoiled all surprise storms.

[uofmiami]

EC 24hr:

[LVblizzard]

Based on the 36 hour map the Euro will make everyone happy.

[tombo82685]

hr 36 sub 996 low due south of hse by 75 miles or soo....h5 low starting to dive into the base of the trof...central and eastern carolinas getting pummeled

[tombo82685]

hr 42 sub 992 about 100 miles east of orf...lgt to mod precip nyc to phl to dc south and east

[tornadojay]

Welcome back to reality euro. Game on

[LVblizzard]

Euro agrees with the GFS. Lock it up!

[tombo82685]

hr 48-54 sub 980 low 100-150 miles off nj shore...phl-nyc hit hard....this site is slow as monkey's

[earthlight]

[Allsnow]

1+ of liquid for phl-nyc on this run

[tombo82685]

hr 60 sub 976 over ack....lgt to mod precip over nyc to phl

[Edge Weather]

EC CAVED TO GFS!!!! 972 off the NJ coast south of Long Island!!!!!!!!!!!!

Already way west of it's 12z run

[weathercoins]

Balt looks ok too Eh?

[Dsnowx53]

I'm almost crying.

[canderson]

Never, really - in fact, as one gets closer to the event the models get better and better, as they have less time for the various errors in assumptions/initial conditions/boundary conditions/data input etc. to propagate out until the actual storm event. Think of it this way: almost every model should give nearly 100% accuracy for forecasted conditions 5 minutes into the future, since the future state is so similar to the current state. Put mathematically, the set of non-linear partial differential equations governing the various physicochemical processes occurring in the atmosphere only has to move forward a small time increment, making the range of possible solutions [actually, numerical solutions are the only way to solve these sets of equations, since analytical solutions are simply impossible] much smaller than for a forecast 24 hours into the future.

This major decrease in the forecast error bars (think of a hurricane's trajectory cone, which gets wider and wider as one moves further into the future) as one gets closer to the event is the reason why people have to put a lot more stock in tonight's 00Z runs than they were able to in the "blizzard" runs from 2 days ago. If the Euro is on board and HPC doesn't come up with any reason to throw out tonight's model runs, then I would think we're clearly ready to lock and load on this event (i.e., significant snows are nearly a sure thing for the DC-Boston corridor; blizzard/snowmageddon conditions are not yet a sure thing) and it would then take a nearly unprecedented total model meltdown scenario, a la March 2001, for this to be a complete miss. Of course, I'm not a meteorologist, so I could be off, but I do know a thing or two about numerical models, physical chemistry, and mass, heat and momentum transfer...

Awesome post. Thank you.