January 2026 OBS and Discussion

[MJO812]

AI models coming around

Model madness.  Not shocking this far out.

[SACRUS]

 

43 / 34 cloudy rain will work its way northeast over the next 4 - 6 hours with the heaviest rains this afternoon into the evening - 0.5 to 1 inch - salt and mush clearer. Dires out Sunday and overall above normal 1/11 - 1/15.  Ridge west balloons pushing trough / colder and potential storms our way in the 1/15 - 1/19 period as much discussed.  Beyond there looks to edge slightly below normal overall the closing week.

 

[North and West]

Honestly I thought someone split LibertyBell down the middle with an axe and krs and geek were the two halves

I hope that LibertyBell is ok!


.

[SACRUS]

Records:

Highs:

EWR: 65 (2016)
NYC: 60 (1876)
LGA: 59 (2016)
JFK: 58 (2016)


Lows:

EWR: 0 (1982)
NYC" -3 (1875)
LGA: 2 (2004)
JFK: 2 (2004)

Historical:

1800 - Savannah, GA, received a foot and a half of snow, and ten inches blanketed Charleston SC. It was the heaviest snowfall of record for the immediate Coastal Plain of the southeastern U.S. (David Ludlum)

 

1800: According to David Ludlum, Savannah, Georgia, received a foot and a half of snow and ten inches blanketed Charleston, SC. It was the heaviest snowfall of record for the immediate Coastal Plain of the southeastern U.S.


1836: “The Big Snow” dumped as much as 40 inches of snow over northern and western Pennsylvania and interior New York State. As much as 2 feet fell across southern New Jersey. 18 inches was recorded at New York City, NY and 15 inches fell at Philadelphia, PA. (Ref. AccWeather Weather History)

1859: The coldest daytime ever experienced in NY City (and throughout New England for that matter) occurred on January 10, 1859. Accurate thermometers were commonplace and well distributed by this time although most of them were not self-registering, meaning that observations had to be made visually, usually three times daily at 7 a.m., 2 p.m., and 9 p.m. The day of January 10 was most likely the coldest ever, based on observations from Montreal, Canada, to New York City. Montreal reported a 7 a.m. reading of -43.6°F on this day, a reading some 15 degrees lower than its modern record minimum of -29°F recorded in 1933. Professor Petty of the University of Vermont in Burlington recorded -31.5°F at 7 a.m. and a bone-chilling 2 p.m. temperature of only -26°F. At Harvard in Cambridge, Massachusetts registered only -4.5°F at 2 p.m. In New York the official thermometer located at Erasmus Hall in Brooklyn registered the following temperature range that day: 7 a.m. -3.7°F, 11 a.m. -7.5°F, 2 p.m. -3.8°F, 9 p.m. -8.0°F. White Plains, just north of the city, reported readings of -13°F at 7 a.m., -10°F at 2 p.m. and -15°F at 9 p.m. This was probably the only day in New York City history when the temperature failed to rise above zero degrees.(Extreme Weather p. 61, by Christopher C. Burt) (Ref. Wilson Wx. History)

1873: A severe blizzard struck Minnesota, killing 70 settlers, injuring countless others and destroying thousands of livestock. (Ref. Wilson Wx. History)

1875: The minimum temperature for the date is -3°F. in Washington, DC. (Ref. Washington Weather Records - KDCA)

1911 - The temperature at Rapid City, SD, plunged 47 degrees in just fifteen minutes. (The Weather Channel)

1949 - Snow was reported at San Diego, CA, for the first and only time since 1882. Snow was noted even on some of the beaches in parts of the Los Angeles metropolitan area. Burbank reported 4.7 inches, and Long Beach and Laguna Beach received one inch of snow. (David Ludlum) (The Weather Channel)

1949: Snow was reported in San Diego, California, for the first time since 1882. The snow was noted on some of the beaches in parts of the Los Angeles metropolitan area.


 

1975 - The "Storm of the Century" hit Minnesota. A severe blizzard moved northward across the state producing up to two feet of snow. High winds drove wind chill readings to 80 degrees below zero, and at Duluth MN the barometric pressure dipped to 28.55 inches. The storm, which claimed 35 lives, occurred on the 102nd anniversary of the infamous "Pioneer Blizzard" in Minnesota. (David Ludlum)
In the warm sector, an F4 tornado struck McComb, MS just after 8 AM, hitting two schools as 325 students took shelter. Fortunately, only 5 people were injured. The twister destroyed 88 homes and 30 businesses across the city. Nine people were killed along the tornadoes 47 mile path. (Ref. AccWeather Weather History)


1981: An arctic blast covered the eastern two-thirds of the country. Wind chills went down to -50° with the coldest wind chill reading of -98° reported at Fargo, ND. Despite the blast only a few locations reported record lows for the date including: Ste. St. Marie, MI: -27°, Providence, RI: 3° and Jackson, KY: 11°. (Ref. AccWeather Weather History)

1982 - The temperature at O'Hare Airport in Chicago, IL, plunged to an all-time record of 26 degrees below zero, and high winds drove the wind chill reading to 77 degrees below zero. The temperaturein Downtown Chicago reached 23 degrees below zero. A week later a second arctic surge sent the temperature plunging back down to 25 degrees below zero. (Weather Channel) (National Weather Summary)

1982: Bitterly cold weather was found across Illinois. Of the 109 weather reporting stations in the state, 48 of those reported lows of 20 degrees below zero or colder. Some of the most frigid temperatures included 27 below zero at Rockford, 26 below zero at Chicago, and 25 below zero at Kankakee and Peoria. The Freezer Bowl was played between the Cincinnati Bengals and the San Deigo Chargers in bitterly cold wind chills. The air temperature was −9 °F, but the wind chill, factoring in a sustained wind of 27 mph,  was −37 °F.

1987 - Bitter cold air invaded the Rocky Mountain Region, with subzero readings reported as far south as Gallop NM. Pocatello ID reported a record low reading of 14 degrees below zero. (National Weather Summary)

1988 - A storm in the northwestern U.S. produced wind gusts to 97 mph at Netarts OR, and up to two feet of snow in the mountains of Oregon. (National Weather Summary) (Storm Data)

1989 - The first documented January tornado of record in Utah struck the south part of Sandy. Asphalt shingles were driven one half inch into a fence. (The Weather Channel) Clear skies, light winds, and up to 24 inches of snow cover, allowed the temperature to plunge to 45 degrees below zero at Roseau MN, and to -43 degrees at Warroad MN. The afternoon high at Grand Forks ND was 16 degrees below zero. (National Weather Summary)

1990 - Strong southerly winds ahead of a Pacific cold front helped temperatures in the central U.S. soar as much as 50 degrees during the day. A total of fifty cities in the central U.S. reported record high temperatures for the date, with afternoon highs in the 70s reported as far north as Nebraska and eastern Colorado. Highs of 53 at devils Lake ND, 76 at Grand Island NE, 73 at Lincoln NE, and 73 at North Platte NE, established records for the month of January. (National Weather Summary)

2001 - Storms dumped more than seven inches of rain on parts of southern California, including Los Angeles, where nearly four inches of rain fell. The heavy rains caused mudslides and road closures along much of California's central coast and up to three feet of snow was deposited in some coastal mountain areas. (NCDC)

[NEG NAO]

1 hour ago, bluewave said:

This is why you have to use the ensemble means which never really showed a big snowstorm threat here. What happens in the age of social media is that someone takes a random outlier model run and uses it to justify that there is going to be a big disruptive snowstorm.

I like a big snowstorm as much as anyone. So it’s very disappointing to turn on YouTube and see multiple videos with over the top big snowstorm titles when the ensemble means have much less dramatic solutions.

Part of this is conditioning from the 2010 to 2018 era when it seems like every time we had a ridge out West and a trough in the East there was a major East Coast KU snowstorm. But that era was defined by a weaker Northern Stream of the Pacific Jet that allowed the Southern Stream to take the lead.

So jet stream dynamics have changed since 2018-2019 making it harder to get big snowstorms over 12” for at least some portion of the area. 

 

I didn't even look at the ensembles  0Z runs because I ran out of time and was spending too much time trying to post maps from the OP runs. Did they show any potential snowfall for a majority of members ? I am satisfied with just a normal snowfall - the 1-3 , 2 -4 , 3-6, 4-8 variety anything over that doesn't grow on trees so to say in the immediate metro so no sense expecting those most of the time. The 06Z GFS for what its off hours run is worth doesn't show the strong wrapped up LP it showed in a previous run - BUT still shows rain changing to snow with a few to several inches Metro wide

[SACRUS]

[North and West]

Nothing to show at all on the 0Z Euro - no snowfall and little precip after this weekend throughout the entire run

I’m flying home then, so I approve this message.


.

[Krs4Lfe]

1 hour ago, bluewave said:

This is why you have to use the ensemble means which never really showed a big snowstorm threat here. What happens in the age of social media is that someone takes a random outlier model run and uses it to justify that there is going to be a big disruptive snowstorm.

I like a big snowstorm as much as anyone. So it’s very disappointing to turn on YouTube and see multiple videos with over the top big snowstorm titles when the ensemble means have much less dramatic solutions.

Part of this is conditioning from the 2010 to 2018 era when it seems like every time we had a ridge out West and a trough in the East there was a major East Coast KU snowstorm. But that era was defined by a weaker Northern Stream of the Pacific Jet that allowed the Southern Stream to take the lead.

So jet stream dynamics have changed since 2018-2019 making it harder to get big snowstorms over 12” for at least some portion of the area. 

 

Yes, and without a return to a KU type storm, it is nearly impossible for the majority of us to reach our seasonal snowfall average because many of us are too warm now for the minor/boundary temperature snow events. So we need to rely on a KU in order to reach our seasonal average. Those have been missing aside from 2021 and January 2022 because of the fast specific flow which has constantly pushed the northern stream out ahead of any southern stream vortex, which has not allowed for amplified solutions. As long as the Pacific jet remains as fast as it is, it will become rarer and rarer to have a KU type storm

[Krs4Lfe]

And verbatim, the GFS model was on an island with an insanely amp solution, and that will not come to fruition. So we have to rely on a strengthening low pressure system that rides the western Atlantic Ridge boundary which can still bring mixed precipitation for us on the coast. Depending on where the boundary sets up after the potential event on the 15th to 16th will dictate the storm track of the next potential event. As we’ve seen since 2020 it is quite possible to have a near perfect benchmark track with the precipitation still falling as rain if the antecedent air mass is still too warm. Just keep in mind that this warm-up is going to last a little longer than expected and will not be finished like the 11th as most models were suggested in a week ago. It will now take into midweek in order for this thaw to finish.

[rclab]

9 hours ago, Cobalt said:

snowman19 only responding in emojis? Why, what happened?

Unless passion overcomes him, his limited posts are well worth reading. As always …..

[Stormlover74]

34 minutes ago, Krs4Lfe said:

Yes, and without a return to a KU type storm, it is nearly impossible for the majority of us to reach our seasonal snowfall average because many of us are too warm now for the minor/boundary temperature snow events. So we need to rely on a KU in order to reach our seasonal average. Those have been missing aside from 2021 and January 2022 because of the fast specific flow which has constantly pushed the northern stream out ahead of any southern stream vortex, which has not allowed for amplified solutions. As long as the Pacific jet remains as fast as it is, it will become rarer and rarer to have a KU type storm

That's just not true at all especially when we're 1/3 of the way there. You don't think we could get 20" in 2 months without a double digit snowfall? 

[Krs4Lfe]

6 minutes ago, Stormlover74 said:

That's just not true at all especially when we're 1/3 of the way there. You don't think we could get 20" in 2 months without a double digit snowfall? 

NYC average is near 30” of snow. It would take several moderate snow events to make it to that level. We have not seen many moderate events in one season since 2015. 2013, 2016, 2017, 2018, 2021 made it above average with storms that had a double digit snowstorm. 2014 and 2015 made it above average with several moderate events (and some bigger ones too, but those were well above average seasons.) would take several moderate events to give NYC another 22” in order to reach average. So unless we buck the trend and get back to having several moderate events like in the great winters of 2014 and 2015, we will need a KU to reach average 

[MJO812]

43 minutes ago, Krs4Lfe said:

And verbatim, the GFS model was on an island with an insanely amp solution, and that will not come to fruition. So we have to rely on a strengthening low pressure system that rides the western Atlantic Ridge boundary which can still bring mixed precipitation for us on the coast. Depending on where the boundary sets up after the potential event on the 15th to 16th will dictate the storm track of the next potential event. As we’ve seen since 2020 it is quite possible to have a near perfect benchmark track with the precipitation still falling as rain if the antecedent air mass is still too warm. Just keep in mind that this warm-up is going to last a little longer than expected and will not be finished like the 11th as most models were suggested in a week ago. It will now take into midweek in order for this thaw to finish.

We heard you for the 100000000 time. 

[donsutherland1]

Central Park, JFK Airport, LaGuardia Airport, and Newark had identical high temperatures yesterday. The last time that occurred was July 26, 2025 when all four locations had highs of 85°.

Yesterday was the fifth time that all four locations had identical highs of 54°. Those dates were:

November 4, 1949
March 23, 1954
February 19, 2011
March 9, 2015
January 9, 2026

 

[Stormlover74]

8 minutes ago, Krs4Lfe said:

NYC average is near 30” of snow. It would take several moderate snow events to make it to that level. We have not seen many moderate events in one season since 2015. 2013, 2016, 2017, 2018, 2021 made it above average with storms that had a double digit snowstorm. 2014 and 2015 made it above average with several moderate events (and some bigger ones too, but those were well above average seasons.) would take several moderate events to give NYC another 22” in order to reach average. So unless we buck the trend and get back to having several moderate events like in the great winters of 2014 and 2015, we will need a KU to reach average 

Correct.. several moderate events 

[NEG NAO]

Anyone else notice that Upton and Mt. Holly in addition to other NWS offices have a new AFD format - using "Key Messages" ??? Hope none of this is AI driven...the computerized voice on their weather radio stations is bad enough

AFD from KOKX

[Stormlover74]

9 minutes ago, NEG NAO said:

Anyone else notice that Upton and Mt. Holly in addition to other NWS offices have a new AFD format - using "Key Messages" ??? Hope none of this is AI driven...the computerized voice on their weather radio stations is bad enough

AFD from KOKX

Yeah they announced it last month I think

[NorthShoreWx]

1 hour ago, Krs4Lfe said:

Yes, and without a return to a KU type storm, it is nearly impossible for the majority of us to reach our seasonal snowfall average because many of us are too warm now for the minor/boundary temperature snow events. So we need to rely on a KU in order to reach our seasonal average. Those have been missing aside from 2021 and January 2022 because of the fast specific flow which has constantly pushed the northern stream out ahead of any southern stream vortex, which has not allowed for amplified solutions. As long as the Pacific jet remains as fast as it is, it will become rarer and rarer to have a KU type storm

Are you a bot?

[MANDA]

35 minutes ago, NEG NAO said:

Anyone else notice that Upton and Mt. Holly in addition to other NWS offices have a new AFD format - using "Key Messages" ??? Hope none of this is AI driven...the computerized voice on their weather radio stations is bad enough

AFD from KOKX

I'll give it a chance....looks like it could be a positive change / addition.  I believe I read most if not all NWS offices will be adopting this new format?

[MANDA]

Rain is looking sparse up this way today.  So far just .04".  Radar clearly shows the bulk of this will be south of my location.  Another precipitation event that is much lighter up this way than what was initially modeled several days ago.  I'm wondering if we'll even break 1/4" up this way and NW of here. 

After today the model consensus AT THE MOMENT looks to be on the dry side for the next 5-7 days with a cooling trend back to seasonable and then colder beyond next weekend.

[Stormlover74]

1 minute ago, MANDA said:

Rain is looking sparse up this way today.  So far just .04".  Radar clearly shows the bulk of this will be south of my location.  Another precipitation event that is much lighter up this way than what was initially modeled several days ago.  I'm wondering if we'll even break 1/4" up this way and NW of here. 

After today the model consensus AT THE MOMENT looks to be on the dry side for the next 5-7 days with a cooling trend back to seasonable and then colder beyond next weekend.

I think the models have mostly been showing a sharp cutoff north of 80. 

[eduggs]

If you look at the ECM, CMC, UK, EPS, GEFS, and GEPS, you think next week has very little chance. We're moving from fantasy land solidly into the mid-range now, and those models don't look very encouraging. But the ECM-AI has improved over the past few runs. And several individual ECM-AI ensembles from 0z also were pretty good hits. That has me at least a little intrigued. Significant snow is still unlikely at this point IMO. Let's start there. But the GFS has shown a negatively tilted trof for multiple runs and we have a second model offering some support. The ECM was also better at 6z than 0z even thought it was not nearly enough. Anomalous trofs are often not well signaled by ensemble modeling.

[MJO812]

Icon was interesting 

[eduggs]

The state of the PNA is loosely correlated to local snowfall in our area. That is a shout out to Don S. But there is no predictive, causal link between the two. Now that we are in range of global weather modeling, the state of the PNA has no bearing on the likelihood of seeing snow next week. We can't predict the exact future state of the PNA any better than we can the exact synoptic evolution. As always, it will come down to minor details in the mid- and upper-level height structure and evolution... as yet unresolved.

[donsutherland1]

This morning's AFD form NWS New York essentially underscores the point made yesterday that at the timeframes involved for potential events (earliest being January 15-16), the guidance cannot reliably resolve the synoptic details. Ultimately, those details are crucial. There's better confidence in the overall longwave pattern, but that's not enough to get to specific forecasting details.

In this context, when using 500 mb vorticity maps, they currently provide insight more along the lines of what could happen vs. what is likely to happen. Over the next 48-72 hours, reliability and model skill will rise to the point where details can be resolved.

Having said this, based on the continuing pattern evolution, my thinking remains that it will be tough to get everything aligned for a significant (6" or above) snowfall Thursday-Friday. Nothing new or significant has emerged to change that general idea. The potential follow-up wave has a better chance of producing measurable and perhaps even a moderate snowfall. More of the 1/10 0z EPS and GEFS individual ensemble members favored the second wave. None of them showed a major snowstorm for either wave. A small handful suggested that the second wave could be significant.

NWS:

Towards Wednesday night into early Thursday the longwave 
pattern amplifies as the ridge out west sharpens, and the trough
in the east amplifies in turn. This amplification is supported 
by the consensus of global NWP and AI guidance. At this time it 
is difficult to time the wave packets, and thus the timing, 
position and strength of storm development. Much of the guidance
is suggesting the frontal boundary slowing and the development 
of low pressure with the amplification of the pattern with the 
upper level flow becoming more or less parallel to the front. 
Suffice it to say there remains a chance of precipitation, with 
precipitation type remaining uncertain during this time frame. 
Therefore impacts are not known at this time.

[eduggs]

4 minutes ago, MJO812 said:

Icon was interesting 

Yeah. Negatively tilted trof with cutoff pretty far south... somewhat like recent GFS runs. I'm less interested in the surface outcome at this point. Gives some support to event potential.

[MJO812]

9 minutes ago, eduggs said:

Yeah. Negatively tilted trof with cutoff pretty far south... somewhat like recent GFS runs. I'm less interested in the surface outcome at this point. Gives some support to event potential.

AI gfs is also more amped 

[eduggs]

A caution flag on relying on ensembles for next week. First I should say that I'm a big fan of ensemble modeling. But there is a theory that they do not perform well with explosive and anomalous trof amplification. Seeing the GFS operational at only 5 days out with its superior resolution have virtually no support from its ensembles for more than one run should raise a caution flag. Add to that the GFS-AI and ECM-AI which tend to dampen waves compared to operational models looking more amplified than the CMC, UK, and ECM. It is possible that a relatively fine-scale shortwave with embedded vorticity might contribute to rapid amplification that is not well resolved by ensemble modeling. I would not be so bold as to predict this. But I think it's something to consider.

[MJO812]

Gfs is going to be better

[MJO812]

2 minutes ago, MJO812 said:

Gfs is going to be better

Way better