Widespread Ice Storm 28-29 December 2025

By Eric Myskowski   May 14, 2026 at 8:00 AM    Email the author
 

1. Introduction:

During the night of December 28 through the day of December 29, a sub-980 mb low passed by to the north and west of the region, with the center taking a track largely parallel to the St. Lawrence Valley but a few hundred miles north. Despite such a powerful low taking a track well to the northwest and a lack of a strong initial cold high to the north, this event caused widespread freezing rain. This was aided by an antecedent cold airmass that extended far south through the Mid-Atlantic, an extensive snowpack, a weak pressure gradient limiting warm-air advection, and a secondary low developing on the triple point to the south. In the end, between two tenths and four tenths of an inch of ice accumulated in most places, though there were a few locally higher amounts just above the 0.5-inch mark east of the Greens and in southern Essex County, NY (Figure 1). Power outages in Vermont peaked at 6,320 during the morning of the 29th and there were widespread travel disruptions throughout the duration of the event.

2. Meteorological Review:

This low-pressure system formed east of the front range of the Rockies, in a typical place for lee side cyclogenesis. It was previously a strong pacific low that came ashore in California that brought several feet to the Sierra Nevadas, with 67 inches reported at Mammoth Mountain. The new low developed on a strong upper trough digging to the west of our region as a ridge tried to amplify to the east (Figure 2). The low quickly strengthened as it moved northeast into the Great Lakes Region, dropping below 980 mb over Michigan. However, this low was running into a modified arctic air mass that had been firmly in place for a few days.

This antecedent arctic airmass provided the surface cold air for freezing rain, and remained in place despite warming air aloft. During the day on the 28th, a strong temperature inversion became established (Figure 3). While the 00Z sounding from Albany, NY (Figure 3) is located just south of the region, it effectively characterizes the regional airmass. By afternoon, Whiteface and Mansfield were in the mid to upper 30s while nearly everywhere else was in the 20s (Figure 4). BTV only hit 27 during the day. With extensive snowpack, light wind, and a December sun angle, the surface was able to radiate very efficiently despite a warming airmass aloft. Of note the National Blend of Models (NBM), the basis of most National Weather Service forecasts, exhibited a pronounced warm bias for both the daily maximums and minimums, failing to resolve the intensity of the arctic air mass.

Figure 4: The 1PM station plot on 12/28/25. Temperatures are shown (black, °F). Not shown are the mid-slopes of Whiteface and Mansfield in the mid to upper 30s. Circled are a selection of high elevation sites. [click to enlarge]

Precipitation arrived during the evening of 12/28 with a typical, elongated warm front. Temperatures were below freezing outside a few parts of the western Adirondacks where localized southerly downsloping was occurring (Figure 5). An elevated warm nose had been firmly in place so any snow was easily melting in the warm layer, and the cold layer beneath it was not deep enough to refreeze the precipitation before it reached the surface. While some sleet was likely present immediately at the onset, the vast majority of the precipitation began as freezing rain.

As is typical in warm-air advection in the region, the temperatures first rose above freezing in the parts of the western Adirondacks and then the immediate Champlain Valley. SLK was the first airport to change to rain, only seeing 0.05 inches of freezing rain before temperatures rose above freezing around 11 PM.  This was modeled well.  By 1:30 AM on December 29th, temperatures at BTV rose above freezing, but not after a quick 0.18 inches of freezing rain accumulated. It took a few more hours for the rest of the Champlain Valley to rise above freezing as many of the areas protected from the winds warmed up slower. By 7 AM, the only below freezing spots were the typical cold areas, east of the Greens, the St. Lawrence Valley and southern Essex County NY. (Figure 6). By around 10 AM, temperatures were above freezing except in the coldest hollows in the aforementioned locations.

A caveat here is that while air temperatures rose above freezing for the vast majority of areas, they did not go up much farther and were several degrees below the NBM almost region wide.  The combination of the barely above freezing temperatures and very cold surfaces from the antecedent airmass caused ice accumulations to continue through the morning after temperatures rose above freezing in many places, particularly on dirt roads. This was even true in parts of the Champlain Valley where temperatures rose above freezing relatively early.  As the low tracked up just north of the St. Lawrence Valley during the day on the December 29th, the pressure fell to its minimum of 978 MB, impressively low for a system that brought relatively light wind on the front side. During the day,  a secondary low quickly began to develop on the triple point that was situated over southern Vermont and New Hampshire, before becoming the primary low over the Maine coast (Figure 7). This development helped further weaken the warm-air advection, and temperatures maxed out in the morning in many places before the final cold front came through in the afternoon.

By the time the occluded front came through, the secondary low matured enough to bring an area of stratiform precipitation back into the region, which intensified over northern Vermont. Under dynamic cooling and cold air advection aloft, the precipitation changed to snow in the mountains, where one to two inches were observed. A few flakes were seen below 1,000 feet, but never in the immediate Champlain Valley. However, the precipitation quickly moved out to the east before more significant accumulations could occur. An interesting phenomenon here is that temperatures more or less cooled from top down, so there was no significant warming from downsloping in eastern Vermont and the Champlain Valley behind the front. By the time surface winds turned westerly, the warmer temperatures aloft had advected out so the associated increased mixing did not cause much warming (Figure 8). To highlight this, Saranac Lake had higher temperatures than Burlington which had higher temperatures than Springfield and Lebanon (Figure 9).

After the passage of the cold front and the associated end of the freezing rain, the effects of the storm system were not over. The pressure gradient behind the low was much stronger than the gradient ahead of it, and strong cold air advection led to efficient mixing. Therefore, there were widespread wind gusts between 25 and 45 mph. However, a southwesterly sting jet moved over the St. Lawrence Valley in the afternoon and was enhanced by channeled flow, leading to locally stronger winds. There, gusts in the 40 to 60 mph range were briefly observed (Figure 10). While a significant amount of ice remained on the trees in the higher elevations east of the Greens and in southern Essex County NY where temperatures barely reached freezing, the winds were not strong enough and the ice was not thick enough to cause a significant increase in power outages. The peak was clearly in the morning from the freezing rain, though there was a minor secondary peak overnight.

QPF Verification - The NBM QPF ended up having a notable high bias for most places. The NDFD forecast for the afternoon of 12/28  mostly utilized the NBM for QPF, though with drier guidance blended in. While the NDFD ended up having a slightly better QPF forecast, it was not by a significant amount and the effects of dropping NBM temperatures  was likely much more impactful on the ice forecast. The caveat here is that this precipitation is for the entire event, not just the QPF that accumulated as ice, and the model QPF as ice was closer to observed than the QPF amount for full event. However, the effects of the too high QPF are still important. 

Ice & Ice-to-Liquid Ratio Verification – The NDFD forecast ice accumulation ratio between 0.6 and 0.7 verified quite well when comparing it to the observed ASOS ratios.  In the end, the ice forecast ended up being slightly high but close (Figure 11). The NDFD had relatively good temperatures even if slightly underforecast, relatively good ice accumulation ratios, but too high QPF.  Therefore, the majority of the forecast error likely came from overestimating the QPF, not from the temperatures or ice accumulation ratios. When comparing the observations to the High-Resolution Ensemble Forecast (HREF) and a strict Freezing Rain Accumulation Model (FRAM) forecast, the forecast ice accumulations were notably overdone, especially in the areas that saw the most ice (Figure 12), and were more overdone than the NDFD in many places.

Acknowledgements: Thanks to Pete Banacos and Robert Haynes for providing feedback and edits.