In the pattern thread I said I was going to start a thread probably tomorrow morning if we get close to a consensus on significant snowfall, but with today's 12Z models showing almost unanimity (but the outlier is the Euro, which is troubling for anyone forecasting snow, as it's still a great model) on at least a major snowstorm Sunday morning through Monday morning, I was thinking of starting it tonight. Some people get their knickers in a twist if I start them too early and then random folks sometimes start them if I wait too long. Oh well. Guess I'll post here as folks didn't seem to like keeping the pattern stuff and event discussion in one thread with creative thread titling (they seem to love it on TOS). Below is what I posted in the pattern thread around 3 pm.
Been out all day playing disc golf and just got back and did a quick check of the models and the GFS, UK, CMC, AIFS, AIGFS, Weathernext2, and ICON all showed a major (8-14") to even historic (14"+) snowstorm for pretty much the entire Philly-NJ-NYC area (and down to DC and up to Boston in most models). However, the Euro shows maybe an inch or so as there is no phasing of the two systems, which the others mostly have.
Historically, the Euro has been the best model (the Euro-AIFS is now, but the Euro is 2nd still), so that is a red flag, but confidence in at least a significant (4-8") snowstorm is growing, although we're still about 5 days from the start of the storm vs. the 12Z model start time (7 am EST and precip as modeled starts around 7 am Sunday i most models +/- 6 hours), so a complete miss is still a decent possibility as the setup, like most NE US snowstorms, is very fragile and sensitive and could still fall apart. There's a thread on this now on AmericanWx...
https://www.americanwx.com/bb/topic/62604-winter-storm-threat-feb-22-24th/page/8/
Summary: Well, Dr. No, King Euro, looks like he might be back, as most of the global models tonight at 0Z caved to the Euro depiction over the last several runs of a much more progressive, flatter storm much further to our SE. So, the chances of a major snowstorm have gone down substantially, but not to zero, and the chances of even a minor to moderate storm have gone down significantly. The AI models, which have done quite well in the 4-6 day range this winter, still show a moderate to significant storm (with only one model the AIGFS still showing a major snowstorm)., while all of the physics-based models show either no snow or only a few inches. The cries you heard last night were from all the snow lovers around the world, lol.
Details: In a stunning, but not completely unexpected development, most of the global models (both AI-based and physics-based) tonight at 0Z caved to the Euro depiction over the last several runs of a much more progressive, flatter east coast trough, resulting from many things upstream, such as a worse ridge out west and poor "wave spacing" between the Fri/Sat system and our Sunday system pieces, such that they don't phase until later, leading to a more progressive storm much further to our SE. So, the chances of a major snowstorm and even a minor to moderate one have gone down substantially, but not to zero, as the AI models, which have done quite well in the 4-6 day range this winter, still show a moderate to significant storm (with only one model the AIGFS still showing a major snowstorm)., while all of the physics-based models show either no snow or a few inches.
The risk of the Euro and many of the ensemble members from ensemble runs yesterday being right in showing minimal snow and things shifting towards a minimal storm are why I wasn't going to start a thread until at least after the 6Z runs this morning (coming out soon) and I wouldn't have started a thread based on what we just saw at 0Z - see my quoted post from yesterday morning below. With regard to the ensembles, at least half of the ensemble members from the Euro, GFS, CMC and AI model ensemble runs yesterday, where the main model is run 20-50 times with initial condition perturbations to probe sensitivity) were showing little to no snow, despite every operational model (the "main" models) other than the Euro showing major snowstorms at 12Z.
Below is a bulleted summary of the differences between the 12Z runs yesterday afternoon and tonight's 0Z runs. As you can see the first 5 (the physics-based models) went from major to huge storms to very little, while the AI models went down, but not as much - they often don't make big moves in short periods like the physics-based models. It's not "over" for a major storm, but the chances probably went from 40-50% to <10% and the chances for even a moderate storm probably went down to 30-40% with a minor to no event now looking like a 50% or more chance. There's still some time left for this to trend back to snowier, but that trend would need to start by Thursday morning at the latest (when the pieces of energy come ashore out west). So, we track, as they say on the weather boards. The NWS discussion is below as is a link to the AmericanWx thread.
- The ICON went from a 12-18" monster for everyone to nada for almost everyone (1-2" for AC to Cape May). That's nuts.
- The GFS went from 6-12+" for everyone at 12Z to 2-5" for most at 0Z
- CMC went from an 18-24" monster for everyone at 12Z to <1" for everyone at 0Z. Also nuts.
- The UK went from 8-14" for everyone at 12Z to an inch along 95 and a few inches for the coast at 0Z. More nuts.
- The Euro had little to no snow for everyone at both 12Z and 0Z. Dr. No.
- The AIGFS went from 12-18" for everyone at 12Z to 6-12" for everyone (the high end at the coast) at 0Z.
- The Euro-AIFS went from 6-12" for everyone at 12Z to 2-3" for 95/3-5" coast at 0Z.
- The Weathernext 2 went from 8-14" for everyone at 12Z to 2-3" for 95/3-5" coast at 0Z.
https://www.americanwx.com/bb/topic/62604-winter-storm-threat-feb-22-24th/page/14/
Area Forecast Discussion
National Weather Service Mount Holly NJ
245 AM EST Wed Feb 18 2026
KEY MESSAGE 3...Monitoring the potential for a winter storm this
weekend, then much colder early next week.
An active pattern looks to continue with much of the model
guidance showing low pressure sliding west to east and exiting
off the northern Mid- Atlantic coast later Sunday. This storm
looks to deepen as it does so, and has the highest potential to
be impactful across our region including the potential for an
accumulating widespread snowfall. Many of the deterministic
models show a storm, and at least some of the ensembles do as
well however they offer some variety regarding the
track/strength of the surface low. The question, as always, is
the track of the surface low. There is still a chance the system
shifts more southeastward which actually has been the general
trend of 00z deterministic guidance. A majority of ensemble and
AI guidance also trended a bit south and east. The amount of
cold air in place ahead of this storm and especially during it
will be important when it comes down to precipitation types and
amounts as well if a storm were to track close enough, as strong
dynamics may also come into play. While it remains way too
early though for specifics, and there likely will be further
trends and changes among model guidance, there remains some
signal for some sort of storm for the second half of the
weekend
The UK, GFS, and CMC all showed big snowstorm hits last night at 0Z, while the AIFS backed off a bit and showed a moderate snowfall and the Euro continued to show no storm at all for us and the AIGFS showed a big storm with mixed precip for 95/coast and snow inland; the ICON showed a miss.
At 6Z, the AIFS went back to showing a major snowstorm and the AIGFS had more snow/less rain, while the GFS showed a pretty big snowstorm, but the biggest news was the Euro finally showing decent snowfall (several inches) for the first time in days and the Euro ensemble showing a nice bump-up in snowfall indicating that the Op move wasn't an outlier. The major ensemble model means all continue to show a good signal for at least moderate to significant snowfall (and Weathernext 2 is similar).
Bottom line is there is still a long way to go, as the storm start (if we get one) would likely be Sunday morning after sunrise, which is about 5 days out. At this point, a moderate to significant snowstorm (with potential for more) is looking like at least a ~50% chance, but a miss to the east is probably still a ~40% possibility (including scenarios where there is some snow at the coast) and a mostly rain event for 95/coast is maybe a 10% probability (rain for all is a very low probability). These are guesses mostly based on the ensemble runs showing probabilities like this and reading what several well-respected mets are saying (I don't do this stuff on my own). IMO, we're not ready for a thread yet, as the miss probability is still too high - if we're still seeing most Op models showing snow in the runs today tonight and tomorrow morning, then maybe a thread tomorrow morning.
