Winter Warmth and the PDO

Blog posts have been few and far between lately, owing to personal circumstances, but hopefully the pace will improve in the near future.  For now I think it's worth highlighting a particularly interesting aspect of this winter in Alaska: the remarkable juxtaposition of far above normal temperatures with a negative PDO phase.

And warm it certainly has been.  While February climate data isn't all available yet, Fairbanks and McGrath both had their 3rd warmest winter (Dec-Feb) on record.  Winter 2000-01 holds the record at both sites, and they differ on the 2nd warmest (1976-77 in Fairbanks, 2017-18 in McGrath), but both came in at #3 this winter.  Here's the time series for Fairbanks:

We (or at least I) tend to think of 1976 as the great climate shift to multi-decadal warmth in Alaska, and that change certainly had a lot to do with the PDO phase shift at the time; here's a paper on that topic:

https://journals.ametsoc.org/view/journals/clim/18/22/jcli3532.1.xml

If we look at Fairbanks DJF temperatures since 1950 with the PDO phase indicated, we see the predominance of a positive PDO in the 1980s, and we also see that virtually all of the very warmest winters have occurred with a positive PDO phase - until this winter:

In terms of correlation, the PDO index has historically shown a moderately strong relationship with Fairbanks winter temperatures, and therefore again winter 2024-25 is a pretty significant outlier:

What seems to be happening is that the PDO-temperature correlation is diminishing over time; it was quite strong before 1990, much less significant in more recent decades, and now it seems this winter has lost the relationship altogether.

There are a few different hypotheses one could pursue on this.  One might be that while the PDO ocean temperature index continue to oscillate (it's just a mathematical construct), the atmospheric circulation patterns that are occurring in tandem with those ocean anomalies have changed in such a way that the PDO now relates differently to Alaska climate.  This winter, for instance, there was an atmospheric ridge over Alaska even though the anomalous warmth in the northwestern North Pacific (and consequent negative PDO) might have been expected to deliver a trough near Alaska.

Another perspective might be that high-latitude warming ("Arctic amplification" of global warming) has become so pronounced that a negative PDO can't be expected to produce historically cool temperatures; even a "cold" pattern is now warm by historical standards.  I think there's some of this going on - broader Arctic temperature trends are partly responsible for this winter's warmth - but in fact the circulation pattern was favorable for warmth in Alaska this winter; it wasn't a "cold" pattern except in terms of the raw PDO index.

It's "interesting" to consider what might happen if all the signals line up on the side of warmth one of these years; it seems a strongly positive PDO index with a classically strong Aleutian low and rampant southerly flow could produce a winter far warmer than any observed in modern climate history.  It may just be that this winter is actually quite "normal" in terms of the climate of the next few decades.

Freezing Rain Trends

Back in December I commented on a significant ice storm in Haines, and I noticed that there seems to have been an increase in frequency of freezing rain events there.  I've been meaning to take a closer look at this.

First, here are some charts showing the number of hours each winter with freezing rain reported by the airport ASOS instruments in Haines, nearby Skagway, Juneau, and also Yakutat (not in the same climate zone but not too far away).  The two winters before this one (2022-23 and 2023-24) both brought a lot of freezing rain to Haines, and 2021-22 saw far above normal freezing rain in Juneau, but unfortunately the data is missing for Haines for most of that winter.

At Yakutat, the data shows a sustained higher frequency of freezing rain since 2016-17; but interestingly this winter is not following suit, with only a couple of brief instances (not included on the chart).

If we average the data from these 4 sites together (using the annual departure from normal at each site), we get an interesting result that does suggest an uptick in frequency since 2021-22.

Does this have any support from the ERA5 reanalysis data?  There is a modest correlation between ERA5 and "ground truth" data for Haines, but surprisingly ERA5 does not show any increase in the last couple of years:

A map view of the ERA5 data confirms no widespread increase in freezing rain across the northern Panhandle region, although an increase shows up in a few places for 2023-24.  Of course the region has extremely complex topography and we would expect huge variations in local conditions; there's no doubt the ~30 km grid spacing of ERA5 is far too coarse to handle this properly, but nevertheless it apparently does have some modest ability to reproduce conditions in Haines, at least prior to 2022.

Here's the average since 2013-14, the beginning of the recent sustained warmth in the North Pacific and Alaska region:

There's no signal for the northern Southeast, so it is difficult to draw a firm conclusion as to whether freezing rain is increasing there or not.

Of course it's a different matter in the southwestern interior, Y-K Delta region, and the northern Bristol Bay coast, where freezing rain has certainly increased in recent years.  Here's a chart of observed frequency from Bethel:

And a longer-term view, bearing in mind that pre-1998 data comes from manual observations, not ASOS, and so there might be (probably is) a bias/discrepancy in how frequently freezing rain is reported.

See this previous post for additional results on winter rain and freezing rain in Alaska and the Arctic:

https://ak-wx.blogspot.com/2022/12/winter-rain.html

Finally, for good measure, here are charts for Anchorage and Fairbanks.  Note that this is for freezing rain, not plain rain (such as occurred last month in many places).

Winter Notes

Here are a couple of quick illustrations of the lack of winter that some parts of Alaska are having.  First, Nome's mean temperature is running at second warmest on record both since November 1 and since December 1.

Remarkably, about a third of days since December 1 have seen a high temperature at or above freezing - including the last 3 days, and 7 of 12 days so far this month!

Down in Anchorage, it's a top-10 warm winter so far, but the lack of snow is the real story.  With only 4.3 inches of snow since December 1, it's the least snowy winter (Dec-Feb) on record so far - although we mustn't forget that there was heavy snow at the very beginning of the season:

https://ak-wx.blogspot.com/2024/11/more-snow-in-anchorage.html

A lengthy spell of daily data is missing from last winter, unfortunately, but we know it was extremely snowy - the second snowiest on winter for the entire cold season.

The following chart shows the remarkable absence of meaningful snow after that early initial dump.  It's quite bizarre.

January Climate Data

Last month was the wettest January statewide in Alaska since at least 1925, according to NOAA's monthly climate data released today.  The previous record holder was January 1949.

Interestingly, none of Alaska's individual climate divisions was record wettest, and that's largely because January 1937 was even more extreme for large parts of the state; but January 1937 was also very dry in (typically wet) southeast Alaska, so it only ranks as (now) 5th wettest statewide.  Compare the two rank maps below:

Here's a map of the difference in absolute precipitation amounts:

As noted in blog posts about the late January storm, some of the precipitation fell as rain as far inland as the Yukon Flats, and with few reliable ground-truth measurements, it's difficult to assess how much snow fell.  No doubt it was "a lot" in areas that dodged the rain and weren't affected by downslope "rain shadow" effects; ERA5 data shows a large area with over 3 times normal liquid equivalent snowfall.  Rick Thoman reports that the highest measured total accumulation was 53" at Wiseman.

The NRCS snowpack map for the end of January shows most of the interior monitored basins in very good shape, but snowpack remained much below normal on the Kenai Peninsula and in Southeast.

ERA5 estimates from a week ago also show widespread ample snowpack for the northern interior and the southern slopes of the Brooks Range, but that's a big contrast from little or no snow for the western Seward Peninsula (if the model is to be believed), the west side of the Alaska Range down to the Alaska Peninsula, and south-central valley areas.

January 2025 was also a very warm month statewide, as we would expect in a wet and cloudy pattern: the correlation between precipitation and temperature is significantly positive in winter, the opposite of summer.  Compared to the last 30 years, the most anomalously warm region was the northeast, where only 2016 was warmer in recent decades.  NOAA/NCEI says that both January 1981 (by some margin) and January 1937 were warmer than January 2016 on the North Slope, although that isn't true at Utqiaġvik.

As for wind, it was a windy month in the vast majority of areas outside of the Panhandle.

The mid-atmosphere flow pattern responsible for all this was anchored by an unusually strong ridge over the northeastern Pacific.  There was also a trough axis over far eastern Russia, and the persistent southwesterly flow between these two features is what transported the warmth and moisture into Alaska.

The position and orientation of the ridge is very similar to what occurred in January 1949, previously the wettest January on record for the state:

Just like in December, the January pattern was very different from the classical La Niña pattern, which involves a trough over western Canada and a ridge over the Aleutians.  Here's the average 500mb height anomaly for 10 years with strong La Niña conditions in January (La Niña is not particularly strong this winter, but this illustrates the point):

With both December and January being top-10 warm months, there's a chance this could be Alaska's warmest winter (Dec-Feb) on record - but only if February ends up in the top 8.  The next week looks more seasonable, so it seems unlikely, but I wouldn't rule it out.

Volatile in the North

First a follow-up note regarding the late January storm.  I was interested in whether the moisture content of the atmosphere reached record levels for January across southern Alaska; I thought it was quite possible based on the prodigious amounts of precipitation.

Looking at ERA5 reanalysis data, however, it's a null result: the precipitable water only set a new January record (1950-present) over a small area near the southern tip of Kodiak Island.  The January record for much of south-central, the southeastern interior, and also northwestern Alaska was set way back in 1963, and on the 16th of that month (Jan 1963) the statewide daily average precipitable water reached 12.4mm (0.49").  For reference, the recent storm's statewide average atmospheric moisture content peaked at 10.0mm.

On another topic, it's quite notable how volatile temperatures have been in northern Alaska in recent weeks.  Umiat has reached or exceeded 32°F on 4 separate occasions in the past 3 weeks, but has also been as cold as -48°F on the last day of January.  And check out the swings in daily maximum temperature at Utqiaġvik since about the 10th of January.

The chart doesn't show it, but the temperature has plummeted again in the last 36 hours: yesterday morning it was only just below freezing in Utqiaġvik, but this evening it's back down to -20°F.  High winds and blowing snow occurred earlier today.

This volatility reflects a profound clash of contrasting air masses: warm subtropical air has been drawn up from the subtropics repeatedly because of a ridge over the northeastern Pacific, but air temperatures aren't much higher than normal overall in the high Arctic.

In other news from Alaska's far north, there's been another subsurface cable break, apparently again caused by ice scouring on the ocean bed.  This happened only about 18 months ago in almost exactly the same location: see this blog post

https://ak-wx.blogspot.com/2023/09/subsea-cable-fixed.html

I find this rather fascinating, because I would certainly not have expected generally thinner and weaker sea ice (than in former decades) to cause severe ice scouring, exceeding the design expectations of the engineers.

Here are a couple of news links about the latest event:

https://www.adn.com/alaska-news/rural-alaska/2025/02/04/beaufort-sea-ice-snips-fiber-optic-cable-limiting-internet-for-some-20000-in-northwest-alaska-until-after-summer/

http://www.nomenugget.com/news/not-nome%E2%80%99s-first-rodeo-dealing-cut-fiber-optic-cable

Storm Notes

Rick Thoman posted a comprehensive write-up of the very warm and stormy weather this month, and I'll add some more notes below.  But first it's worth noting the dramatic change since the weekend: this morning's cold was impressive across the western and northwestern interior.  The temperature dropped below -50°F at quite a number of locations generally near the Koyukuk River valley, and even Selawik in the northwest reached -49°F.  Click to enlarge:

Looking back at the weekend warmth, Fairbanks was continuously above freezing from 7am on Friday through 8pm on Sunday, i.e. 61 hours, and this is certainly the longest such stretch recorded in January - only one other day (January 15, 1981) had a daily minimum temperature above freezing.  No such days have been recorded in February, and the only ones in December occurred during the great chinook/thaw of early December 1934.  The recent warmth can't quite compare to that historic event for overall warmth (highs in the 50s back then), but there are no other events in the depths of winter that have produced the duration of thaw seen in Fairbanks over the weekend.

The historic nature of the warmth was also illustrated by 43°F at Tanana, a new record for the month of January (120 years of data!), and 49°F at Tok.  The situation at Tok was truly record-shattering, with a low of 39°F on Sunday; this means the daily mean temperature was 44°F, a full 6°F warmer than any other day in December through February, with data back to the 1950s.

The sustained flow of moisture-laden air from the southwest produced prodigious precipitation in many places.  The map below shows estimated 7-day totals in terms of liquid equivalent.  The SNOTEL instrument at McGrath reported 3.8" of precipitation, which far exceeds the record for 3-day precipitation total in winter (2.10" in 1990, data back to 1939).  Near Talkeetna, 3-day totals of 4-5" were measured at relatively low elevation, and of course much more in the mountains.

Up on the southern slopes of the Brooks Range, snowfall was very heavy - in Wiseman, for example, where snow depth is just about at a record level for the time of year.

Here's a map of reported snow depth today, courtesy of the NWS:

One more aspect that I noticed: as cold moved into northwestern Alaska on Sunday, and Fairbanks remained in a thaw, the difference in daily mean temperature between Fairbanks and Kotzebue reached +46°F, and this is also a record for the time of year.  Typically of course Fairbanks is colder in January, and in the past such a large positive temperature difference has only occurred on rare occasions in late winter through spring.

We'll be able to do more analysis on the weekend's historical context when the ERA5 data is available in a couple more days (I'm particularly interested in the moisture content/precipitable water).

More Extremes

No time for details tonight, but a couple of quick notes regarding the ongoing extreme weather situation for Alaska.  First, it was southern Alaska's turn for high winds today: gusting over 60 mph in Anchorage and Palmer, 81 mph at Delta Junction (the highest in 25 years), and 90 mph out west (sustained at over 80 mph) at Cape Romanzof.

Regarding temperatures: Fairbanks spent its third straight day in the 30s Fahrenheit, and a quick look at the data suggests this is an unprecedented warm spell (at least since 1930) for the month of January.  The great thaw of early December 1934 was greater, and late February 1943 had 3 straight days with a mean temperature of 35°F or higher, but in January Fairbanks hasn't even had 2 straight days with a mean temperature of 35°F or higher.  There were only 6 individual such days from 1930 through 2024 - but Fairbanks has just seen three in a row (assuming the temperature doesn't drop off sharply before midnight tonight).

More commentary and wrap-up on this historic event in the coming days.