January Forecast Discussion

[mahk_webstah]

I just think that type of solution won't verify, til the Euro jumps on board I'm skeptical, as bad as its been this year.  Many of its ensembles are amplified but it has plenty of flat solutions too.  The NavGEM rule would suggest this event may be far from settled...remember, often when its fairly close to other globals at Day 5-6 often times the globals make a major shift...its fairly rare to see a NavGEM idea at 120 plus hours verify and right now its damn close to everything but the Euro

Wasn't your thinking a couple days ago was that this would not cut and would not be amplified.  I think some others had the same feeling given the fast gradient between the SE ridge and the cold air to the north.  Are you getting a different feel for this scenario now?  Seems evidence for a more amped system is building as it gets closer.  Would love to see a somewhat less amped system.

 

Also isn't the rule that when the Nogaps (which tends to be suppressed) shows close to the coast then the storm is probably staying close to the coast if that is what is being shown by other models.  At least that is how I have heard it used in the past.

[SnowGoose69]

Wasn't your thinking a couple days ago was that this would not cut and would not be amplified.  I think some others had the same feeling given the fast gradient between the SE ridge and the cold air to the north.  Are you getting a different feel for this scenario now?  Seems evidence for a more amped system is building as it gets closer.  Would love to see a somewhat less amped system.

 

Also isn't the rule that when the Nogaps (which tends to be suppressed) shows close to the coast then the storm is probably staying close to the coast if that is what is being shown by other models.  At least that is how I have heard it used in the past.

 

I did think it would probably end up flatter and also felt if it did go the route most 00z runs showed we would not see much frozen precip at all nor would inland areas because the positioning of the high was poor and the air mass was stale.  The 00Z runs definitely show that with the Euro showing hardly any frozen precip near the coast.  At the same time I said yesterday as did another poster we're better off taking our chances with it being more amplified since there would be some chance of snow albeit not a great one. The NavGEM rule applies more inside 4 days,  beyond 4 days there have been some people who've noticed that when it agrees with other globals closely those other models can frequently make a significant shift in the next day or two because the NavGEM does not hit too many ideas accurately that far out...so there may still be more surprises coming with this.

[mahk_webstah]

I did think it would probably end up flatter and also felt if it did go the route most 00z runs showed we would not see much frozen precip at all nor would inland areas because the positioning of the high was poor and the air mass was stale.  The 00Z runs definitely show that with the Euro showing hardly any frozen precip near the coast.  At the same time I said yesterday as did another poster we're better off taking our chances with it being more amplified since there would be some chance of snow albeit not a great one. The NavGEM rule applies more inside 4 days,  beyond 4 days there have been some people who've noticed that when it agrees with other globals closely those other models can frequently make a significant shift in the next day or two because the NavGEM does not hit too many ideas accurately that far out...so there may still be more surprises coming with this.

Thanks for the response.

 

Basically for me the GFS is a good storm, the Euro a snow to ice adv/warning event.  Positioning of the high not so disastorous up here as it is in SNE.

 

Interesting about the Nogaps - didn't know that.  Obviously there is something injested in the data that pushed most models well NW of where they were before, be interesting to see what happens in this next run.  12 will show if this is a new trend, a blip, or something else.  

[SnowGoose69]

Thanks for the response.

 

Basically for me the GFS is a good storm, the Euro a snow to ice adv/warning event.  Positioning of the high not so disastorous up here as it is in SNE.

 

Interesting about the Nogaps - didn't know that.  Obviously there is something injested in the data that pushed most models well NW of where they were before, be interesting to see what happens in this next run.  12 will show if this is a new trend, a blip, or something else.  

 

The Euro has just been bad, it did not appear to necessarily be its bias either, when you look at the 12Z run yesterday its 96 hour frame was ejecting the low similar to the GFS out of the southwest...the Euro simply was night and day with the energy diving out of Canada phasing in and thats the key with this storm...that energy is over Siberia right now, it gets over AK in 48 hours roughly, at that point we could see another model swing because that piece is what is needed for this storm to go way NW, if that does not hook up with the southern energy this is likely flat and maybe too flat...we may want to see a dampening of both disturbances to result in some phasing but nothing too big.

[donsutherland1]

This morning, the Arctic Oscillation (AO) rose to +0.553. That represented a rise of 1.329 standard deviations from yesterday's figure. The latest GFS ensemble forecast suggests that the AO could approach +3.000 in coming days.

 

 

Although the EPO- has increased prospects for a colder than normal January, and a possible Arctic intrusion sometime after January 5 could increase those prospects, the positive AO is not an encouraging development for excessive snowfalls. First, 85% of January KU-type snowstorms commenced with an AO-/PNA+ pattern and 92% commenced with an AO-. 56% of 6" or greater snowstorms in January also commenced with an AO-/PNA+ pattern and 78% commenced with an AO-.

 

Of course, a strongly positive AO in early January does not preclude a KU snowstorm later in the month or afterward. For example, January 2005 started with the first 10 days having an AO above +2.000. 8 of those 10 days were +3.000, including 3 that saw the AO exceed +4.000. January 5 had an AO reading of +4.703. After than, the polar vortex began to break down. By January 21, strong blocking began to develop. Shortly thereafter, a blizzard brought heavy snow and high winds to parts of the northern Mid-Atlantic and New England regions. February and March featured abundant blocking. January 1983 and January 1993 provide additional examples. 1983 featured a blockbuster February blizzard. 1993 featured the March "Storm of the Century."

 

Winter 2004-05 also featured strikingly similar ENSO conditions to the current winter. For example, on December 22, 2004, the ENSO Region 1+2, 3, 3.4, and 4 anomalies were respectively -0.2°C, +0.6°C, +0.7°C, and +1.1°C. On December 24, 2014 (the latest available figures), those readings were -0.2°C, +0.7°C, +0.7°C, and +0.9°C.

 

It should be noted that numerous winters that saw January start out with strongly positive readings did not feature a 2005-type flip. Nevertheless, given that the EPO has gone strongly negative (as it also did in late December 2004) unlike a number of those more lackluster years and strikingly similar ENSO conditions, there arguably remains hope for a turnaround.

 

For now, at least it is getting colder.

[Monmouth_County_Jacpot]

I think it is fairly obvious now that this will not be like last winter.

it's been obvious since thanksgiving don't know why people are so shocked about it now

[Yanksfan]

This morning, the Arctic Oscillation (AO) rose to +0.553. That represented a rise of 1.329 standard deviations from yesterday's figure. The latest GFS ensemble forecast suggests that the AO could approach +3.000 in coming days.

 

 

Although the EPO- has increased prospects for a colder than normal January, and a possible Arctic intrusion sometime after January 5 could increase those prospects, the positive AO is not an encouraging development for excessive snowfalls. First, 85% of January KU-type snowstorms commenced with an AO-/PNA+ pattern and 92% commenced with an AO-. 56% of 6" or greater snowstorms in January also commenced with an AO-/PNA+ pattern and 78% commenced with an AO-.

 

Of course, a strongly positive AO in early January does not preclude a KU snowstorm later in the month or afterward. For example, January 2005 started with the first 10 days having an AO above +2.000. 8 of those 10 days were +3.000, including 3 that saw the AO exceed +4.000. January 5 had an AO reading of +4.703. After than, the polar vortex began to break down. By January 21, strong blocking began to develop. Shortly thereafter, a blizzard brought heavy snow and high winds to parts of the northern Mid-Atlantic and New England regions. February and March featured abundant blocking. January 1983 and January 1993 provide additional examples. 1983 featured a blockbuster February blizzard. 1993 featured the March "Storm of the Century."

 

Winter 2004-05 also featured strikingly similar ENSO conditions to the current winter. For example, on December 22, 2004, the ENSO Region 1+2, 3, 3.4, and 4 anomalies were respectively -0.2°C, +0.6°C, +0.7°C, and +1.1°C. On December 24, 2014 (the latest available figures), those readings were -0.2°C, +0.7°C, +0.7°C, and +0.9°C.

 

It should be noted that numerous winters that saw January start out with strongly positive readings did not feature a 2005-type flip. Nevertheless, given that the EPO has gone strongly negative (as it also did in late December 2004) unlike a number of those more lackluster years and strikingly similar ENSO conditions, there arguably remains hope for a turnaround.

 

For now, at least it is getting colder.

With the extreme snowfall in Eurasia this year, many mets said that would correspond with a -AO. Having a +AO instead would be unprecedented, and in my book would forever change my thinking at long range forecasting.  

[IsentropicLift]

12z GFS OP through 96hrs - The northern stream is getting involved quicker. The SE ridge is also a tad stronger this run becuase the flow out West is more amplified.

[IsentropicLift]

Surface low heading towards the Ohio Valley. Some buckling in the isobars near the Carolina coast.

[IsentropicLift]

Hour 108 some rain/snow/sleet mix to start, surface freezing line hanging just NW of 95.

[IsentropicLift]

This is a pretty nasty run for the Poconos and places like Sullivan County, NY. It's a very good run however once you get north of say the CT/MA border.

[IsentropicLift]

Transfering to a secondary off the NJ coast. Quite the ice storm for parts of the LHV.

[IsentropicLift]

Ends up 1008mb just northeast of the benchmark. QPF is 0.75-1.00" for all locations in this sub forum.

[IsentropicLift]

12z PGFS has more phasing than the OP. It's almost a complete phase. It's a much deeper system. 992mb surface low over Michigan and 980mb in Canada. We end up much, much warmer and wetter with very heavy rain overhead. Almost 1.75" of rain in six hours. The surface freezing line ends up in Central Maine. 

[Doorman]

http://www.tropicaltidbits.com/analysis/models/gfsp/2014123012/

 

gfs- para combo

 

 

[donsutherland1]

With the extreme snowfall in Eurasia this year, many mets said that would correspond with a -AO. Having a +AO instead would be unprecedented, and in my book would forever change my thinking at long range forecasting.  

Long-range forecasting is still a very difficult endeavor with not a great deal of skill. Unfortunately, should this winter have a predominantly positive AO with little or no blocking, that outcome will probably suggest that the early promise associated with the SAI may have been due to a combination of a still modest sample size, additional factors that drive the AO, etc. This will be a disappointing outcome, as the SAI seemed to offer promise with respect to one of the major teleconnections that shape the winter pattern. However, consecutive years of underperformance with the measure would suggest that one can't overlook the possibility that the index isn't as robust or reliable as it initially seemed. Right now, though, it's still premature to assume that the AO won't become predominantly negative down the road, even if the current ensemble forecasts are anything but promising.

[IsentropicLift]

Well the 12z PGFS ends up looking a lot like the 00z ECWMF ensembles.

 

We still have a few things going for us.

 

1) We're still over four days out and whether or not the system cuts and how deep it becomes is heavily reliant on northern stream energy that is still over a very poorly sampled area in Siberia.

 

2) Models typically struggle at this range with the extent of low level cold air. So if the modeling is too strong with the parent low or if secondary development does occur and we can salvage the mid-levels, things might work out. Right now I would go with a mix to rain for the coast and snow to a mix to rain for the interior. And unfortunatly I think parts of the interior LHV could be in for quite a nasty time of it, especially if the parent low ends up weaker.

 

On Saturday there is still plently of cold air around, so if things come in quicker than forecasted, that could make a difference as well as the high would be in a more favorable position.

[BoulderWX]

Well the 12z PGFS ends up looking a lot like the 00z ECWMF ensembles.

 

We still have a few things going for us.

 

1) We're still over four days out and whether or not the system cuts and how deep it becomes is heavily reliant on northern stream energy that is still over a very poorly sampled area in Siberia.

 

2) Models typically struggle at this range with the extent of low level cold air. So if the modeling is too strong with the parent low or if secondary development does occur and we can salvage the mid-levels, things might work out. Right now I would go with a mix to rain for the coast and snow to a mix to rain for the interior. And unfortunatly I think parts of the interior LHV could be in for quite a nasty time of it, especially if the parent low ends up weaker.

 

On Saturday there is still plently of cold air around, so if things come in quicker than forecasted, that could make a difference as well as the high would be in a more favorable position.

only place I can see accumulating snow with this one would be LHV on north - could be some slop to start N/W of the city to start but would be very surprised to see anything accumulate for anyone in NJ/NYC/LI

[uncle W]

the last typhoon in the Pacific didn't recurve like the one in October...was that a hint of a lousy pattern for us?...I should of gone to Florida for January...I probably wouldn't have missed much...

[BoulderWX]

the last typhoon in the Pacific didn't recurve like the one in October...was that a hint of a lousy pattern for us?...I should of gone to Florida for January...I probably wouldn't have missed anything...

fixed it for you 

[Animal]

only place I can see accumulating snow with this one would be LHV on north - could be some slop to start N/W of the city to start but would be very surprised to see anything accumulate for anyone in NJ/NYC/LI

 

Agree. fooking disaster of a run.

More rain in January.

[Ralph Wiggum]

12z PGFS has more phasing than the OP. It's almost a complete phase. It's a much deeper system. 992mb surface low over Michigan and 980mb in Canada. We end up much, much warmer and wetter with very heavy rain overhead. Almost 1.75" of rain in six hours. The surface freezing line ends up in Central Maine.

And there's what you wished for....taking your chance with more northern stream interaction.

[IsentropicLift]

And there's what you wished for....taking your chance with more northern stream interaction.

I've got no problem with a rainstorm or a cutter. I don't get all bent out of shape on December 30th becuase things haven't gone my way.

[IsentropicLift]

only place I can see accumulating snow with this one would be LHV on north - could be some slop to start N/W of the city to start but would be very surprised to see anything accumulate for anyone in NJ/NYC/LI

I'm not really concerned with accumulating snows at this point. That ship sailed days ago. It still has the potential to be a very impactful event in spots.

[IsentropicLift]

Here is a possible red flag, the 12z GFS-ENS-2P5 which is the ensemble mean for the PGFS, only deepens the low to 1016mb and sends it more towards Toronto instead of Detroit. That could be a sign that this should be less amplified like the GFS OP shows.

[donsutherland1]

the skinny

on gfs para implementation

 

http://www.nws.noaa.gov/om/notification/tin14-46gfs_cca.htm

If the data for the last 31 days is representative, I am somewhat uneasy about implementation of the Parallel GFS on 1/14. In most of the categories against which anomaly correlations are measured, the existing GFS has modestly outperformed the parallel version, both for the 0z and 12z cycles. For the others, the performance has been about the same. I'd prefer to see the parallel consistently outperforming the GFS before the change takes place.

 

Below are the latest anomaly correlations between the two versions (120 hours for the Northern Hemisphere):

 

In terms of root mean square error for 120-hour heights, the statistics are as follows:

 

0z:

GFS: 44.28 (better)

PRX: 45.85

 

12z:

GFS: 45.88

PRX: 43.51 (better)

 

In short, I don't think one can state that the Parallel Version is qualitatively superior to the version that it is replacing, at least not based on the last 31 days of data.

 

Source of Verification Statistics:

0z: http://www.emc.ncep.noaa.gov/gmb/STATS_vsdb/

12: http://www.emc.ncep.noaa.gov/gmb/STATS_vsdb/12Z/

[SnowGoose69]

The 12Z GFS if it verified as shown is not terrible for the coast...there would be snow at the start and likely more than the GFS is showing, as is always the case with these events the models miss the overrunning a bit and also warm things up below 850 too quickly outside of 60-72 hours...of course there is little margin for error, a solution more amped than the GFS forget it, a solution less amped we're in businnes for something similar to the early December event last year.  One thing that changed in the last 3-4 days is that the system was delayed enough ejecting that it another arctic high is now able to come sliding across the northern tier because the orginal one was going to be very out of position if this ejecting 36 hours earlier.

[IsentropicLift]

As warm as the 12z GGEM is, even that model has a lot of ice for parts of the interior. Typcially in SWFE the precipitation arrives sooner than forecasted, keep that in mind.

 

[IsentropicLift]

The Euro is going to have a lot more northern stream interaction this run. Coming in a lot more amplfied at 96hrs and the PAC Jet is much further SW.

[IsentropicLift]

The southern stream energy is also ejecting faster out of the southwest so that's causing it to run out ahead of the energy coming out of the northwest some despite being more amped up.