Serious Cold

I'm on the road and away from my usual data resources at present, but a quick post is in order to highlight the remarkable National Weather Service forecast for the next few days in Fairbanks:

This morning's NWS forecast for the international airport calls for daily high temperatures below -40° from today through Sunday, i.e. 5 days in row.  This hasn't happened since the intense New Year's cold snap at the turn of the century (Dec 30, 1999 through Jan 3, 2000).

The anomaly is even more striking in comparison to the last 15 years: there have only been 2 individual days with a high temperature of -40° or below at the airport since January 2009.  Early January 2009 saw 4 straight days with a high of -40° or -41°F.

The Fairbanks record for consecutive days below -40°F is 9 days in December 1961.

And here's the forecast for Fort Yukon: chilling out below -50°F for the next several days.

Looking farther ahead, there's a distinct glimmer of hope for significant relief in the second week of January:

Coldest yet for Fairbanks

The latest round of severe cold has set up a bit farther west, with the lowest temperatures today over central Alaska rather than far eastern Alaska and the Yukon Territory.  Most of the Fairbanks area dropped well into the -40s, and Eielson AFB saw -53°F.  Ice fog formed in the valley, as illustrated by this panoramic webcam view from the university's West Ridge (click to enlarge):

It will be interesting to see if the temperature dips to -50°F at the airport before a modest warming trend ensues.  The negative half-century mark has become quite rare in the official climate record from Fairbanks:

There's little doubt the decline in the most severe cold is attributable both to the local urban warming trend and the background (regional/global) warming trend.  Other, much less developed interior sites also show a decline, but less overwhelming.

Sea Ice Update

The extreme cold over southern Yukon is making headlines at the moment - and deservedly so, given how persistent it is - but for balanced perspective it's also worth considering the remarkable warmth in the Arctic in recent months.  Rick Thoman notes in his latest post that Arctic sea ice extent is running at a record low for the time of year, as it has been since late November:

https://alaskaclimate.substack.com/p/winter-solstice-2025-arctic-sea-ice

Zack Labe's splendid graphic highlights the anomaly:

The largest deficits are found to the east and north of Svalbard, where ice-free waters extend to nearly 85°N, according to the NSIDC analysis.  This is pretty remarkable in late December:

It might be counter-intuitive given the preponderance of cold in Alaska, but Bering Sea ice is also below normal, mostly because of ice deficits in the central and western Bering Sea.  There is pretty good ice cover to the east of St Lawrence Island:

The warm anomaly in the Arctic has been ongoing since the height of summer.  Of course most recent years have been warmer than earlier decades, but the latter half of 2025 has been particularly unusual:

Extended Cold

My posts this month are taking on a particular theme - cold - as the ongoing cold spell is starting to make this winter feel like an "old fashioned" La Niña winter for eastern and southeastern parts of the state.  (See this post for comments on cold spells in La Niña versus El Niño winters.)

The cold has been most intense for the southeastern interior, where Northway is having its coldest December to-date since 1980, and that's with the first 4 days of the month being very warm.  The contrast with the very persistently and unusually warm autumn is striking:

Remarkably, Northway saw -50°F on 4 consecutive days from the 8th through the 11th, and 1980 was also the last time that happened before the middle of the month.  The only other Decembers this century with -50°F in Northway at any time in the month were 2012 and 2022.

The state's lowest measurement this month came from Chicken, of course: -56°F on the 9th.  With data back to 1997, that's the second earliest date for such cold; the earliest was December 1, 2012.  So far this month, the average temperature in Chicken is below -32°F, and that's on track for the coldest calendar month on record (again, during the short period of record).  There have only been three Januarys with an average temperature below -30°F in Chicken: 2004, 2012, and 2020.

It's really only the far eastern interior that has been extremely cold, as the core of the cold air has been in northwestern Canada, transported southward on the east side of a very strong ridge over the Bering Sea.  The two following figures show the average 850mb temperature departure from normal and the average 500mb height since December 5, when it turned cold:

It's interesting to note that the persistence of this pattern - and it shows no real signs of stopping - was successfully anticipated by long-range forecast models.  My post on December 6 highlighted this, and the latest forecast update shows much the same signal (above-normal 500mb heights near the Date Line) for several more weeks (refer to the bottom of the post for an explanation of this graphic):

If the model (in this case the ECMWF subseasonal model) is correct, the cold will shift westward slightly in the coming weeks, implying that Fairbanks may not have seen the worst of it yet.  But of course the model could well be wrong.

Note on interpreting these figures: the format takes a bit of getting used to.  The idea is to show the progression of anomalies (departures from normal) from top to bottom in time, with longitude varying from left to right; refer to the map cutout at the bottom for a visual reference of longitude.  The values on the chart are averages from 50-80°N.  The top part of the figure (labeled as "ANALYSIS" on the right) is the progression that has already happened, in this case the 30 days prior to December 18.  The lower part of the figure (labeled as "FORECAST" on the right) is the predicted progression from the ECMWF 46-day model, extending out to February 1st in this case.

The last figure shows temperature anomalies and indicates that below-normal temperatures are expected to persist well into January over Alaska and western Canada, and there's a hint of westward progression in the core of the cold.  The previous figure shows 500mb height anomalies (i.e. pressure in the middle atmosphere), and the persistent orange shading near 180°W is the Bering Sea ridge.

A Cold Snow

Friday's snowstorm in Fairbanks was unusual from at least a couple of standpoints.  First, the total of 10.5 inches is a hefty fall in Fairbanks; more winters than not never see a snowfall this heavy in the span or one or two days, and sometimes it doesn't happen for several years.  For instance, from winter 1997-98 through 2007-08, it happened only once (January 2000).

Still more unusual was the low temperature during such a heavy snowstorm.  Nearly all of the snow fell with a temperature below 0°F, and the high temperature of 13°F didn't occur until late Friday when the snow was ending and a westerly breeze picked up and mixed out the low-level inversion.  Here's the balloon sounding from Fairbanks airport at 3pm on Friday, when the substantial snow was just ending: notice the sharp low-level inversion, with surface temperatures around -20°C (-4°F).

Typically a heavy snowstorm is much warmer, because the strong flow required to transport and dump all that moisture also mixes the warmer air aloft down to the surface.

Remarkably, the heaviest rate of snowfall occurred in Fairbanks on Friday morning when the temperature was -11°F.  Looking back at the full history of observations from Fairbanks, only one year in the past saw moderate or heavy snow reported at a colder temperature: it occurred twice in 1955 (-14°F in January and -16°F in December).  It is simply very rare to have heavy snowfall rates at these temperatures in Fairbanks.

Looking at calendar days with more than 10 inches of snow in Fairbanks (of which there are only 20 in the last 95 years!), Friday was the coldest such day on record.  The previous coldest high temperature was 15°F on February 11, 1966, and December 29, 2016.  Friday's daily mean temperature of -1°F was also the lowest daily mean temperature for such a snowy day, the previous record being +1°F.

Why so cold?  It seems the explanation lies with the fast-moving nature of the disturbance and the fact that it occurred without any warm, southerly component to the mid-atmosphere flow.  There was no fundamental change in the cold pattern, with a strong ridge hanging out over the southern Bering Sea throughout the entire event, and the mid-atmosphere flow merely went from northerly to westerly and then back to northerly.  Here are 500mb maps from early Thursday, Friday, and Saturday mornings:

The surprise to me is just how much snow fell with such a fast-moving westerly disturbance; I would have expected something closer to 3 or 4 inches.  But this was no quick burst of snow from a fast-moving front; visibility was 1 mile or less for more than 8 hours.  An unusual event, indeed.

Wind and Cold

First, the cold: there have been a couple of reports of -50°F in the ongoing cold snap across the eastern interior, according to the NWS.  The Birch Creek RAWS to the east of Circle Hot Springs reached the negative half-century mark yesterday evening, and the Robertson River co-op observer near Tanacross also measured -50°F this morning.

Also of note, Northway airport made it to -47°F this morning, colder than anything observed last winter.

Down in south-central Alaska, high winds hammered exposed locations throughout the weekend and were still causing big trouble today.  The Palmer airport illustrates the extreme wind episode, with wind speeds gusting over 50mph more or less continuously since midnight on Friday night.

Here's a news article:

https://www.adn.com/alaska-news/mat-su/2025/12/07/emergency-shelter-opens-in-mat-su-as-thousands-remain-without-power-amid-hurricane-force-winds/

The peak gust of 85mph on Saturday morning is the highest credible wind gust reported from Palmer, with data back to 1973.  [Update: Jim Green correctly notes that the "peak wind" section of the METAR actually reported higher gusts, as high as 100mph on Sunday morning.]  Of course the usual caveats apply about changes in wind measurement instrumentation over the decades, and for the chart below I'm only showing results since 1997, when the ASOS platform was commissioned in Palmer.  Also, the anemometer was upgraded to a sonic instrument in September 2006.

Prior to this weekend, the highest wind speed was observed in early January 2022, and that wind storm was also driven by a very strong north-south pressure gradient.  Compare Saturday afternoon's MSLP and 500mb maps (below) with those from the 2022 event, as documented in the following blog post:

https://ak-wx.blogspot.com/2022/01/mat-su-wind-storm.html

In Saturday's post, I commented that the MSLP difference between Deadhorse and Seward was approaching an all-time record.  According to ECMWF AIFS data (very similar to ERA5 reanalysis), the Deadhorse-Seward difference reached 59mb on Saturday afternoon, and the peak 24-hour average exceeded 56mb.  Only one event in the ERA5 history (1950-present) had a higher 24-hour mean pressure difference: February 2-3, 1970.

This really has been an extreme event.

Cold Outbreak

The first serious cold spell of the winter is now under way for Alaska, as strong northerly flow has brought cold air plunging down from the Arctic in the past day or so.  Temperatures have dropped below -40° at the usual cold spots, and Fairbanks airport is sitting around -30°F this morning for the first time since last January.

The immediate cause of the cold outbreak is a powerful high pressure ridge that ballooned up over the Bering Sea and far eastern Russia in the past couple of days, driving a sharp change to northerly flow over Alaska.  The shift in the flow pattern has been very dramatic: check out the 500mb flow change over Alaska from Thursday morning to this morning:

The north-south pressure gradient across Alaska is extreme today, with MSLP near 1040mb on the Beaufort Sea coast and near 980mb on the south coast of the Kenai Peninsula.  As a result, strong winds are producing ground blizzard conditions and nasty wind chill in exposed locations.

Remarkably, today's near-60mb north-south pressure difference is close to (or may be) an all-time record; here's a chart showing the annual maximum and minimum MSLP difference between Deadhorse and Seward, based on daily mean data.

The ERA5 record for the daily mean pressure difference is 55.8 mb on December 8, 1955.  I need to rerun the analysis using hourly data and compare it to today's conditions.  Note that very strong pressure differences like this are found in winter, when the jet stream is strongest and flow anomalies are greatest.

Looking ahead, there are strong signs that the Bering Ridge setup could have some serious staying power, judging from model forecasts.  For example, here's a diagram of 500mb heights averaged over the 50-80°N latitude band, with the time progression from top to bottom.  Look at the yellow/orange colors at the left and right sides in the forecast section (below the dashed line): the ECMWF model expects a ridge to persist near the Date Line all the way into mid-January.

This would mean persistent, significantly below-normal temperatures for southeastern Alaska and western Canada.  The CPC forecast is locked into this signal for their 6-10 day, 8-14 day, and week 3-4 forecasts:

Declining Alaska Observations

Yesterday UAF's ACCAP announced the results of a study evaluating the availability of weather data from FAA and NWS observing sites in Alaska:

https://uaf-accap.org/projects/alaska-weather-station-data-availability/

Unfortunately the results are not pretty: there has been a serious decline in availability of hourly weather data in recent years.  The number of sites reporting consistently on an hourly basis has dropped by about 40% in the last decade, and only 19 reliable FAA/NWS sites remain in off-road-network locations.  Check out the linked article for some striking graphics.

It's not difficult to observe the trend in the data.  Here's a chart for Utqiaġvik showing the percentage of days each year with temperature and wind data from at least 18 of 24 hours (not necessarily at the top of the hour):

Last year the observation reliability fell of a cliff, with virtually the entire summer missing data in the overnight period.  This year the same thing happened in March and April, but not in summer.

From a scientific standpoint it's a shame to see the lack of commitment to maintaining weather and climate monitoring with the FAA/NWS instrumentation, and from a public service standpoint it's bad news, as reliable real-time ground-level weather data is important for forecasting and decision making.

A significant counterpoint, however, is that NOAA's Climate Reference Network has expanded its coverage across Alaska in the last 20 years.  There are now 25 sites installed, and while several aren't reporting currently, the overall volume of data has become substantial in recent years.  The chart below shows the annual number of daily observations with maximum and minimum temperature across all the sites:

With the CRN sites being located preferentially in stable and often remote locations, and with high-quality instrumentation, the scientific value for long-term climate monitoring is potentially much higher than with airport instruments; but for real-time weather monitoring and forecasting, the CRN network can't replace the loss of FAA/NWS data in recent years.

Stratospheric Disruption

Meteorologists around the Northern Hemisphere have been excitedly watching events high up in the stratosphere recently, as the circulation aloft has become very unusual for the time of year.

Normally at this point on the calendar there is a robust vortex of cold air circulating high above the Arctic, but in recent days this vortex has been pushed well off its axis by a big ridge over the Canadian side of the Arctic.  The anomaly is ongoing; here's a forecast pressure map for Friday at the 10mb level, or about 100,000 feet above sea level.

The average 10mb westerly wind at 60°N is hovering around zero, which is approximately the weakest it's ever been at this time of year; more typical values are 20-40 m/s (45-90 mph).  The green line in the chart below shows the recent evolution of this closely-watched "U10/60" index, and the red line shows a forecast.

What does this have to do with Alaska, you may ask?  Well, the high pressure anomaly aloft also extends down through the lower stratosphere and the troposphere, so it's connected to the surface-level weather pattern over Alaska.  Here are maps for Friday at 100mb (about 50,000 feet) and 500mb (18,000 feet):

As of this morning, the 500mb ridge is centered over the Chukchi Sea, bringing relatively clear skies and cool northeasterly flow to northern and western Alaska:

Forecasters like to watch the stratospheric winds because they tend to foreshadow circulation patterns at lower levels, sometimes giving valuable information weeks in advance of important pattern changes.  In other words, what happens in the stratosphere doesn't stay in the stratosphere.

For instance, here's an analysis of 500mb heights 10-40 days following previous dates this century when the 10mb 60°N westerly wind dropped to zero in winter.  Above-normal heights near Greenland in the map below signify a negative Arctic Oscillation, a well-known effect of such "sudden stratospheric warming" events.

Closer to home in Alaska, there tends to be a trough over northern Alaska and a ridge over the Gulf of Alaska, and that is usually a mild setup for southern Alaska:

More strikingly, there is a strong wet signal for western Alaska: liquid-equivalent precipitation is commonly above normal in the month or so after these major stratospheric disruptions.

It will be interesting to see if the historical pattern plays out this time around.  CPC does show a strong wet signal in the 6-10 day period, but focused on the eastern interior at this time:

Autumn Clouds

Following up briefly on last week's post, I took a closer look at cloud cover and solar radiation data from the Fairbanks area, seeking to document the remarkable lack of sunshine in the last few months.  All lines of evidence indicate that it has, indeed, been persistently and unusually cloudy in Fairbanks since mid-August.  Here's a chart showing the daily solar radiation measured since summer at the high-quality CRN site just to the northeast of town:

Of course it's typical for rain and cloud to increase across much of Alaska in August (see this post), but the change this year was more pronounced than usual for the interior, and the weather remained generally cloudier than normal all the way through autumn.

For the 90 day period ending November 10, the CRN instrument measured less radiation than in any other year, with data going back to 2002.  Interestingly, the last 3 years have seen the lowest solar energy in this short history, so it's a persistent anomaly that I strongly suspect is related to the remarkable North Pacific warmth of the last few years.

ERA5 reanalysis data provides a much longer history of estimated solar radiation for the same window, and it sends a similar message:

Of course most of the solar radiation in this autumn window occurs in the first month or so, while the sun still has some strength, so it's worth looking at cloud observations from the airport ASOS for a more evenly distributed perspective on cloudiness.  Here's the result: this was the cloudiest autumn since at least the mid-1950s, if the cloud cover observations are to be believed (and there are certainly inconsistencies in the observing method over time).

To get a sense of the spatial extent of the recurring anomaly, here are solar radiation anomaly maps for the August-October window back to 2020, the last relatively sunny autumn in Fairbanks.

Again, the analysis for the whole seasonal period is focused on the earlier part of the window (mainly August) because of the annual cycle.  However, the last 5 years have also been cloudier than normal in each month separately, so it's not just a late summer change.  The maps below also show a similar trend in the Arctic - at least according to the ERA5 model data.  It's not good news for autumn aurora watchers, or for those who enjoy clear and crisp weather at that time of year.