Most of December and the first two weeks of January were very cold in most of Alaska and northwest Canada. I posted about the cold in the first half of December here, the cold snap through Christmas here, December overall for Alaska here and Arctic context for the cold snap in the December climate review here.

This post serves as an event recap, though because of the breath and duration of event this review is necessarily selective. Also I have only limited tools to assess the historical context of this event in northwest Canada, so I’ve surely missed some important highlights. Because this event was international scope, I provide temperature information in both degrees F and degrees C. This makes the post longer, but I hope it’s relevant whatever your preferred temperature scale.

The 2025-26 cold snap was distinguished primarily by the long duration of the very cold weather rather than daily extremes. Based on ERA5 Land reanalysis, for Alaska overall, every day but two between December 4 and January 15 was colder than the 1991-2020 baseline average (Fig. 1 top), making this the most significant prolonged cold snap since 2012 (which was colder). For the Yukon Territory, every day between December 5 and January 9 was colder than average (Fig. 1 bottom).

The coldest period for northwest North America overall was December 5 to January 15 (Fig. 2). The departures are quite markable given this is a seven week period.

For the Yukon Territory overall the coldest day was December 22, with temperatures south of 65°N widely 20°C or more below normal (Fig. 3). The very low temperatures extended into the eastern Interior, but Alaska west of about 160°W was notably warmer than normal.

The coldest day from the central Yukon Territory westward into eastern and central Interior Alaska was January 4th (Fig. 4). Temperatures in southern YT were not quite as low as around winter solstice. With above normal temperatures across the Alaska Peninsula this wasn’t quite coldest day for Alaska overall.

The coldest day for Alaska overall was January 8, anchored by the deep cold over southwest mainland but temperatures widely 15°F (9°C) or more below normal over mainland Alaska (Fig. 5). At this point temperatures had moderated considerably in the Yukon Territory and Southeast Alaska.

The lowest temperatures recorded during December and January (Fig. 6) were notable, though not to record levels across Interior Alaska and the Yukon Territory. Both Tok and Chicken, Alaska, with lows of -63F (-52.8C), recorded the lowest temperatures at their respective locations since January 2009. Fairbanks’ and Tanana’s lowest temperature was the same as the low in 2024, and Bettles had a lower temperature last winter. Carmacks, YT reported the lowest temperature since 1996 but Whitehorse, Dawson, Beaver Creek and Pelly Crossing all had the lowest temperature only since 2022.

Beyond the Yukon and Interior Alaska there were a few daily records and other short term extremes during this cold snap.

Juneau airport low temperature of -10F (-23.3C) on December 22 and 23 were daily record lows and also the lowest temperature there since 1995.

In Southcentral Alaska there were a scattering of daily record lows the first and second weeks of January. Talkeetna set a daily record of -38F (-38.9C) on January 8. This is the first daily record low to be set in January since 1975. Also notable was the exceptionally cold January 3rd at Portage Glacier Visitor Center southeast of Girdwood. Calm winds and clear skies allowed the temperature to stay low all day, with a high temperature of -22F (-30.0C) and a low of -30F (-34.9C) the coldest day at this location since the observation site was established in 1998, though January 1989 probably had at least one day as cold or colder. Kenai’s low temperature of -36F (-37.8C) on January 8 wasn’t a daily record but it was the lowest temperature there since 1999.

In Southwest Alaska, King Salmon set daily record lows on two days and Bethel on one day.

The duration of cold snap is illustrated in these site-specific highlights:

• Fairbanks Airport:

  • 21 days with lows of -40F/C or lower, all between December 14 and January 13, the most since the winter of 1970-71

  • 57 hours with dense ice fog (visibility one-quarter mile or lower), the most since the winter of 2008-09

  • 30-day average temperature -31.0F (-35.0C) December 15 to January 13, the lowest since 1970-71

• Tok: 22 days with lows of -50F (-45.6C) or lower, all between December 8 and January 6

• Chicken: 29 days with lows of -40F/C or lower, all between December 6 and January 10

Notable “consecutive days” streaks include:

• Fairbanks Airport:

  • 32 straight days high temperature 0F (-17.8C) or lower (December 14 to January 14), second longest on record

  • 43 straight days with daily low temperature -10F (-23.3C) or lower (December 4 to January 15), second longest on record

• Tanana:

• Eagle (cooperative station):

  • 42 straight days with daily high temperature 0F (-17.8C) or lower, (December 5 to January 15), longest on record

• Northway:

• Anchorage:

• Dawson, YT: 19 straight days with lows -40F/C or lower, longest since 1996

The daily temperature ranges are shown at Fairbanks and Tok (Fig. 7). Fairbanks, the westernmost location shown, had two notable moderating events in December, both associated with pulses of warmer air aloft moving from the Bering Sea into the central Interior and both produced significant snowfall. The first of these made as far east as Tok. Also notable at all these valley locations are a number of very cold days with only a few degrees spread between the high and low temperature. These days were clear (excepting local ice fog) and calm and vividly illustrates that although the sun scraps above the horizon for a few hours, in December and early January this provides no significant solar heating poleward of 60°N.

The cold was particularly unrelenting at Dawson, YT in December (Fig. 8, top), although the coldest day occurred on January 4. At Whitehorse (Fig. 8, bottom), there we two distinct episodes of deep cold, one in early December and another around winter solstice.

An historical perspective on this cold snap at Fairbanks is shown in the times series plotting the coldest 10-day periods each winter since the early 1900s (Fig. 9). By this measure, this cold snap was slightly colder than 2011-12 and not quite as cold as 1988-89. However, about a dozen winters prior to 1980 had a 10-day period colder than this event.

The mid-atmospheric flow pattern and its slow evolution in December and January explains a lot of what happened on the ground. During the second half of December (Fig. 10, left) extremely strong high pressure aloft over the Bering Sea supported high pressure at the surface over eastern Interior Alaska and the Yukon. Because of the lack of solar heating, clear skies allowed continuous escape of heat from the top of the snowpack to outer space, with the cold air pooling in valleys. In early January (Fig 10, right) the pattern changed, with a general westward shift of the high and low pressure centers aloft compared to December. For example, the the Being high pressure shifted northwest to be near the Sea Of Okhotsk, while the low pressure over central Arctic Canada in December shifted west to be over the Beaufort Sea in early January. This westward “retrogression”, so-called because it’s in contrast to the more usual west-to-east progression of mid-latitude weather features near ground, is a common feature of the mid-atmosphere flow pattern during the cold season.