The below is part of the first official Premium forecast for the upcoming Saturday winter storm threat. To view the forecast in full, including snowfall accumulations and timing of the snow, and to have all future storm updates summarized and emailed directly to you, subscribe here. Otherwise, below we break down the most recent model trends for the upcoming storm:
Well, the latest model guidance is now out for the day, and it has done little to show what really to expect with the upcoming storm as a wide number of scenarios continue to remain on the table. In this post, I will again run through a large number of weather model guidance solutions for the upcoming storm to show what remains on the table. Thus far, the models have been moving in the way I generally expected with slight further south trends sparing SWCT/NY most of the worst impacts. This is what I outlined the last two days, and as will be seen a number of models have almost no impacts around and north of I-84, with only limited impacts to the south. Still, a slight jog to the north and we are seeing a blizzard along the coastline, and I do not want to underemphasize the strength of the storm if there were a northern jog (despite it not appearing likely at this time). With that, let’s dive into the latest models and see what remains on the table for this storm.
Let’s begin with this image of the 12z GFS ensemble members for 7 PM on Saturday. It is interesting for a number of reasons. First, clearly the storm has been delayed, as heavy snow is shown lingering across parts of the region into Saturday evening. That has again been a major model trend today to be slower with storm development and precipitation tracking up to the north. The top left was the operational GFS, which showed a decent hit bringing anywhere from 6-14 inches of snow across the region (higher amounts near the coast, of course). We see most members agreed with that, but a few have the storm slightly further south and east, while few have it more north and west. Perhaps most interestingly, we have two models that show a warm layer working in around the 850mb during the storm. This could turn some snowfall across the region to sleet as mild ocean air is spun onshore by the storm. The GFS suite has been showing this potential the most, but other short-range guidance like the NAM has also been showing the potential for snow to turn/mix with sleet, which could also limit accumulations across some of the region. Still, most models keep us all snow, and that is the favored solution. But even with that, there remains an incredible spread in individual solutions.
Here is the output for 7 AM on Saturday among the members, where we see most show snow just beginning across the region. This should limit impacts to the day on Saturday, as no longer does the Friday evening commute look at risk from this storm. That is a relatively high-confidence observation. Now here we have the GFS ensemble member accumulated precipitation through 1 PM on Sunday. Using this, I was able to correctly identify the trend the last couple days on the GFS and other operational weather models. So what do we see now? The bias appears still to be slightly to the southeast from the operational GFS, but not quite as much. 2 members of the 12 have the storm entirely miss the region, which remains a possibility but appears unlikely at this time. The general consensus among the GFS ensembles appears to be for .8-1.1 inches of liquid at the coast and .3-.8 inches of liquid inland. This would roughly translate to 6-12 inches at the coast and 3-8 inches inland, though that will still need some adjustments that I will go through below. The main story, however, is just how sharp a gradient the ensemble members are predicting, as is almost every major weather model. Central New Jersey should very easily see 1-2 inches of liquid, bringing up to 1-2 feet of snow. Yet northern CT may not see any liquid and escape the storm unscathed, again because of how early the upper level low cuts off which allows the surface low pressure center to strengthen significantly earlier and not move up the coast as much. When looking at the GFS ensembles, we see that a northern shift could put us in the danger zone, but also that the bias appears to still be slightly to the south and east which means that SWCT/NY could easily avoid the worst.
Fitting into this trend, the 18z GFS mirrored my above thoughts almost perfectly, as seen by the 4 AM Sunday output here. On the bottom one over from the left, we can see forecast total liquid from the storm. Parts of Long Island are shown having up to 2 inches of QPF, or close to 2 feet of snow. Yet I-84 sees barely a quarter of an inch of QPF, bringing with it just 2-4 inches of snow. Closer to the coast we see around half an inch to maybe three quarters of an inch of liquid, but this is certainly not the 1-3 feet that some places farther south will be seeing. Southern Westchester looks to have the best chance of seeing really impressive snowfall amounts as they will be closest to the storm. Still, it is simply too early to know exactly where this gradient will set up, and we could literally be looking at a range from 3 inches to 2 feet in a matter of about 50 miles. This will be one of the strongest gradients I have ever seen, and means that up until 24-48 hours before the storm there could be very high uncertainty with this storm unless forecasts shift much more significantly south from here. Even on the GFS with less snow, coastal areas could see wind gusts to 30-40 mph, and gusty winds are almost a given, but inland areas would not even see many wind issues at all as the strongest winds are across New Jersey and further south.
While I hate using the NAM in its long-range as it is a very inaccurate model outside of 48 hours, I want to show this image for Sunday at 1 AM. Focusing in on the bottom right, we show that the NAM is predicting an even sharper snowfall gradient. 2 feet of snow is shown falling in New York City, with Danbury CT seeing maybe an inch of snow and Greenwich CT seeing around a foot. The exact placement of this gradient will shift any number of times as models try to figure out the exact track of this storm, but hopefully this demonstrates just how tricky the upcoming forecast is, and how parts of the SWCT/NY region could go from seeing very few impacts to almost blizzard conditions in one forecast update, as a jog of just 20-30 miles could really change almost everything.
Making this more tricky is that most other international weather model guidance also agrees on this incredibly sharp gradient setting up. Canadian guidance herethrough 7 PM on Saturday shows over a foot of snow in New York City but only a few inches in Danbury, similar to what the GFS was showing. The operational ECMWF shows almost 2 feet in Staten Island but only around 6-8 inches in Greenwich and 2-4 inches in Danbury. These subtle forecasts shifts will create amazingly different impacts across the region, and it is just too early to say with high confidence exactly where this gradient will set up, though recently the 18z GFS ensembles indicated it could set up so far south that even the coast sees minimal snowfall accumulations. Overnight weather model guidance should add a significant amount of clarity, so for now we sit and wait to see what changes come and adapt the forecast accordingly.
To begin receiving more detailed updates regularly emailed to you, and to view the first official Premium forecast alongside future school and travel impact analysis moving forward, subscribe here. The first official Free forecast will be published tomorrow evening around 7 PM outlining expected snowfall accumulation and timing as well.