Roger Smith

With the new year approaching, I am terminating the 2025 contest thread with the results of Dec 2025 and the annual contest, and moving all new activity over to this thread where you can post Jan 2026 forecasts at any time, as is our custom in January, no late penalties will fall until later into the first week, let's say for sure no penalties before end of Jan 2nd, so take your time and post / edit freely -- if you don't see a table of forecasts you can edit because I only see the forecasts when I make up the table. For anyone new wanting to enter, the forecast contest asks you to predict temperature anomalies at nine locations across the U.S., in F deg relative to 1991-2020 averages. The nine locations are DCA _ NYC _ BOS _ ORD _ ATL _ IAH _ DEN _ PHX _ SEA ... Happy new year ...

Yes, Edmonton Alberta is running 12 F below normal (7 C below) but that colder regime has stayed locked into central Canada most of the last week to ten days. It hasn't even been below average here near the US border. Just a general note to all participants, I will open up a thread for 2026 contest forecasts and see how that goes, perhaps we will continue on, and perhaps not. Depends on level of interest, it's easy enough for me to keep doing what I normally do but at the same time, I don't want people to enter out of anything other than an enthusiasm for tackling the challenge. So if anyone happens to post Jan 2026 forecasts after this post, I will move them over to the new thread which will be open almost as soon as I post this. Happy new year, I am not going to post any preliminary scoring for December, we will let the results be a total surprise on January 1st or 2nd.

I didn't say I agreed with it, I said it was an operational error. It definitely shows zero snowfall on Long Island. But that's not my forecast.

snowdepth change graphic, never shows any accum on Long Island. I need to make it clear that I understand how to read weather models and I consider this an operational error in the GFS model, not valid guidance. I have made no forecast assumptions after seeing this erroneous graphic. Thank you for your attention to this matter (is that how it goes?)

06z GFS may have an operational error there showing no snow over all of Long Island while precip panels maintain previous spread of shield, either that or the model now thinks it would fall as freezing rain. The way the cold air has settled, can't see this being realistic, 7 to 10 inch snows quite possible everywhere but within 10 inches of CPK ruling device. Local 10-14 possible at higher elevations n NJ and se NY.

Looks like rapid spread into PA around noon EST, could be some freezing drizzle earlier than that. Maybe if you left very early (like 0400) you could make it to destination before the storm cranks up. This storm will tend to explode over a wide area rather than moving gradually, because it needs to saturate lower levels to get started. On the GFS hourly precip panels, it covers almost none of PA at noon and most of PA at 2 p.m.

Locally heavy snow potential Toronto to Hamilton ON Friday with lake-enhanced bands 10-20 cm possibly more around Oakville-Burlington.

Newspapers don't speak weatherweenie language, they mean best case scenario 10" there. I'll say 5-10 inches sw CT and 3-5 w MA into e CT, 1-3 ORH-PVD and Tr-1 BOS, se MA.

This might be the first time dew points above 60 in KY are recorded on a day before a winter storm hits NYC but those dews are not even going to reach se VA, if the GFS scenario is correct the low that forms takes all day to cross Ohio and is slowly consumed by the attempt to push the cold air back, eventually it just reforms off the southern Delmarva and continues east from there. In this kind of setup I don't see a lot of potential for guidance to bust on the mild side, I think there might be a chance it busts on the cold side further south and sleet mixes for a while in c NJ then it goes back to all snow there too. NYC will be in the vicinity of 534-540 dm thickness during the bulk of the event, often that is rather marginal for temps but I note that right now, the air mass pushing southwest has an average T/Td of 20/10 under that portion of its outflow (approx upper MI across L Huron into sw ON) and the surface feed is even colder from upstate NY so by morning I would not be surprised to find temps in n NJ and NYC metro near or a little below 20 F. And there will be quite limited warming before precip arrives. I bet there will be numerous comments posted tomorrow just before onset along the lines of "it is much colder than I was expecting" and the result will be precip boundaries further southwest than a lot of the guidance is showing. So in other words the bust will be opposite to what a lot of people are expecting the bust to be. Would be comfortable with 5-8 inches generally and 10-15 local banding maxima. There again those bands could be intense further south than some are expecting.

What Friday commute? Santa going back to North Pole?

Widespread 60+ dewpoints in developing storm's warm sector, as long as it turns southeast far enough west to avoid flooding low to mid levels in NYC region with 40-45 deg air, I could see this becoming a heavy snow situation for wherever surface temps are closest to 28 F. And that currently looks like ne PA to n/c NJ to NYC and LI. It is a WAA event except that it is going to be rotated around by 45 deg compared to many of them, which I believe will tighten up gradients and lead to a wall of snow outcome. There will be a lot of thunder and lightning across PA into s NJ and n DE during this storm's development phase Friday afternoon-evening. Key point to evolution is that a weak wave tracks e.s.e. across region tonight, reinforcing surface cold and that colder air will flood into the east coast region before the low really reaches full potential, so that will create a fairly resistant cold layer below about 850 mbs. Taint factor may be overestimated for NYC metro by some, but time will tell.

06z GFS not going for the north trend and looks reasonably similar to previous GFS runs.

and it was 70 F two days later. In this case, more like 55 F to 60 F?

That NAM fiasco is due to the low running 200 miles further north, it's north of Erie PA on the map instead of somewhere near the WV-OH border into sw PA. It's over the left guardrail into the Walmart parking lot.

I think your outcome is mainly sleet and freezing rain bands with snow northeast of Baltimore possibly, but there is probably a one in ten chance this becomes your snowstorm at the last moment, the low is definitely going to have considerable energy from the high dew points it will encounter over the mid-section of the country, while right now it is just dealing with dry air masses over the Rockies (freezing fog in Miles City MT is about all it is producing so far). When it gets into Iowa and Illinois it will have a lot more moisture available and bands of sleet and freezing rain will develop around s WI and MI, then snow will develop over s ON and nw PA, w NY. If this turns hard right over w PA you could get a semi-surprise snowfall into DC and most of MD. Looking at all guidance the solutions are further apart than usual at 36h-48h lead time. But if I had to choose I think the GFS is probably nearest to reality at this point and its snow axis is ne PA to n/c NJ. Thundersleet around mid-day Friday is possible for s PA into ne MD.

Reggie slid into the left guardrail and some of the other guidance is up against the right fence, but GFS sailing down the actual intended path, this model run reminds me of rush hour in Salt Lake City (stay behind a truck in the middle lane and you'll survive it perhaps) ... I believe so anyway. Low is now in n.e. WY and air mass cold enough to snow (with just freezing fog) is in Montana and North Dakota. It's sunny and mid-70s in e CO and most of KS, but foggy and low 50s in s IA into central IL, this is the moisture available to this storm basically, so it should begin to drop moderate amounts of sleet and freezing rain in WI and MI tonight into Christmas Day, transition to snow will start over ne OH, w NY and s ON Christmas night into early 26th. I can't see it missing south, too much development taking place before the transfer to a C Hat coastal low. Axis of heaviest snow seems likely to be ne PA into n/c NJ. Northeast winds off LI Sound should help with banding and accumulations.

A more general answer about outcomes for heavy snowfall Decembers in the past ... 40 Decembers have seen 8.0" or more (of 157 incl 2025 so far) 29 Decembers have seen 10.0" or more. There were only four seasonal totals below 30 inches in these forty winters, and only one below 20 inches. All of those were well back in the record, the more recent snowy Decembers have generally surpassed 40 inches in seasonal totals. Of course, the snowier the December, the less extra snow is required to break any given target, but as a general rule it appears that you can triple to quaadruple total December snowfall to get a reasonable estimate of most winter totals. I noticed a typo in the NWS discussion, they say 8-10 inches is a worst case scenario, we all know they meant best case scenario. (merry Christmas) I will stick to my first call of 3 to 6 inches for the region in general and 8 to 12 inch local banding near Raritan Bay into areas like New Brunswick and Somerville NJ.

I suspect there will be a convergence zone near the 28F isotherm in n/c NJ with enhanced totals similar to that event last winter (?) (or was it two winters back now?) ... 3 to 6 inches might be a good general forecast with this narrow band of 8 to 12 inches possibly somewhere like Long Branch - Sayreville-NB-Somerville NJ. That convergence zone would represent the southern boundary of unmodified arctic air, a second boundary further south would limit the mixed precip zone across parts of central NJ and e PA. That one would likely run through n DE and ne MD into the PA/MD border region. These two frontal zones would both likely stall for several hours then drift south near end of event, so mixing north of the strong convergence zone might be negligible.

1947 was 26.4" (26.1" on 26th) and 2010 was 20.0" (12.2" 26th + 7.8" 27th). That 7.8" portion is the daily record for the 27th. The 12.2" on 26th would have been a daily record on every day of December except the 15th (12.7" fell in 1916), the 20th (15.5" fell in 1948), but even without 1947, not the 26th .... 18 inches fell in 1872 on Dec 26th during a very cold spell. Boxing Day seems good for snow, I think Don showed it was the most likely date in December to see measurable snow, and other significant falls include 7.0" in 1890, and 11.2" in 1933. In 1969 there was 6.3" over the 25th and 26th. The three largest December totals for two days all include the 26th.

2 inches for CPK, I'll believe that when I see 0.5" posted in the climate summary. My take on this one is, light coatings of snow grains, mixed with drizzle, not a great road scenario because it won't slow down traffic very much while adding the hazards of slushy partly frozen crud. Could drop 1-3 inches on colder northern suburban areas though. Widespread black ice the following morning.

Warm ridge in central US is slowly retrogressing and this Friday (possibly into Saturday as per CMC) wave is sliding down the retreating forward edge, a good scenario as it will lock in cold on northeast side of the wave. Also there's a good 50/50 type low forming as the Tuesday minor event explodes into a deep low south of NS and heads for s.e. Newfoundland. That locks in the cold high. I could see this being an over-performer and giving the NYC region 8 to 12 inches of snow. Mixing issues will be way south, like PHL to DCA. I expect the timing may begin to shift a bit towards Friday overnight into Saturday morning. Looks like it is followed by a slight warming trend before extreme cold descends after the arctic vortex amplifies over the Lakes. That could lead to another 1-2 inches of snow from outer edges of rapidly developing coastal near Cape Cod. This is an oddball if not unique pattern setting up so analogues won't be very helpful.

Probably the way models are trending, it's becoming a question of whether the milder air ever makes it past central NJ but I think temps will spike at around 45F for a few hours late Christmas Day, then slide down into high 20s setting up a solid wall of icy resistance when the Friday wave drops southeast, could lead to a much heavier snowfall than forecasts are currently saying (can't say record because it's 78 years after 12-26-1947). A few days ago it was looking like a very mild two days 24th-25th and now it's looking more like a couple of hours of slight warming. Encouraging when the models bust on the warm side.

I would bet when it warms up slightly for Christmas Eve into the 25th, it will get very foggy until the warm sector is suppressed south. Foggy and 48-52 F temps, Santa will not find every house. Oh right, I'm supposed to pretend Santa doesn't exist. Forgot about that.

Tracking anomalies and projected end of month ... _____________________________ DCA _NYC _BOS __ORD _ATL _IAH ___DEN _PHX _SEA _____ (anom 1-16) __________ -9.2 _ -8.7 _ -7.7 __-10.7 _-4.9 _-3.2 __+8.3 _ +5.1 _ +6.4 _____ (p anom 1-31) ________ -5.0 _ -4.5 _ -4.0 __-6.0 _-2.5 _-1.0 __+6.0 _ +3.5 _ +2.5 (17th) _ A more variable regime will cut into large negative anomalies in eastern and central regions. It will stay generally mild in the western regions but SEA will have a larger reversal as arctic air builds in for a while later in the month.

Strong winds also up here, with resultant power failure after trees came down on power lines. Near blizzard conditions in southern Alberta, major highway multi-vehicle pile up north of Calgary. We have lost our thin snow cover during recent mild spell, quite cold in the wind but it cleared up shortly after a couple of brief flurries, still quite windy, 35 F.