Warming Up Fast

A very strong upper-level ridge over Southeast Alaska and northwestern Canada has produced a big warm-up for eastern and southern Alaska in the last week, with temperatures reaching record levels for the time of year in some spots.  Yakutat reached 64°F on Friday, which would be the warmest on record this early in the season if it were not for a similar warm-up in 2021.  Numerous locations set daily record highs on Friday and/or Saturday, from Ketchikan (65°F) in the far south to Bettles in the northern interior (56°F).

Here's the 500mb analysis from 4pm AKDT on Thursday, courtesy of Environment Canada:

The most unusual warmth I spotted was at Cordova, where the daily minimum temperature of 47°F on Thursday was the highest overnight minimum this early in the season by a full 3 weeks.  Bettles also saw a record warm night for this early, with a low temperature of 39°F on Thursday despite having 29" of snow on the ground at that time.

As of the last daily report, Fairbanks snow depth is down to 3", so the official meltout date may well be today.  The ice at Nenana is looking very rotten, and the accumulation of thawing heat units is now in the historical range for breakup.  However, with unusually thick ice this year - measured at 50" a couple of weeks ago - perhaps it will be a few more days before the tripod goes out.

With meltout occurring earlier this year than the last two seasons (both in early May), green-up and fire season in the eastern interior will probably come to life a bit earlier as well.  Here's an article discussing this, with a few comments from Rick Thoman:

https://www.webcenterfairbanks.com/2024/04/17/fires-floods-risk-with-early-may-green-up/

It makes sense that there would be a connection between early snowpack loss and increased early-season fire activity, so I had a quick look to see if this shows up in the data.  The chart below shows June 1 fire acreage statewide versus the snowpack meltout date in Fairbanks:

There does seem to be something of a relationship, although it's marginal and the sample size is small (I only have daily fire data back to 1995).  If we step forward to July 1, the relationship seems to disappear:

The highest July 1 acreage on the chart was in 2022, when the Yukon-Kuskokwim region saw unprecedented early fire activity owing to a very dry spring; there's a lot more to the puzzle than just the timing of meltout.

https://ak-wx.blogspot.com/2022/06/early-fire-and-may-climate-data.html

Breakup Outlook

Breakup concerns will be front and center in many minds over the next few weeks.  The National Weather Service is assessing a somewhat elevated risk of flooding in some locations along the Yukon and Kuskokwim Rivers, as illustrated by the following map:

Generally above-normal snowpack is the main concern, along with some risk of below-normal temperatures in the latter part of April, raising the possibility of a more dynamic breakup with ice jams.  However, the temperature outlook has warmed in the last few days for central and eastern Alaska, so it may be western regions that are most at risk.

Here's the breakup outlook:

https://www.weather.gov/aprfc/breakupProducts

Looking at the Tanana breakup at Nenana, there's still some way to go before the tripod is at risk; it's warm today (nearly 50°F), but the accumulation of thawing degree days hasn't made much progress yet, as is typical for the date.  As of yesterday, Fairbanks had reached 20 TDDs (accumulation of daily mean temperatures above freezing), and at least 75-100 are needed for breakup - and typically more like 100-150.

It's interesting to see some multidecadal variability in the amount of thawing needed for breakup at Nenana.  This could be attributable to variations in ice thickness and snowpack, both of which affect the breakup date.

Despite this variability, however, simply using the accumulation of TDDs allows the breakup date to be predicted with a correlation of 0.9; it's mostly about temperature.  Here's what the (in-sample) predictions look like using each year's date when TDDs reached the historical median for breakup:

Notice that very late breakups occur earlier than predicted, because the sun is an increasingly important factor as May advances.

Obviously the TDD calculation uses 32°F as the threshold for accumulating heat units, and I got to wondering if a different threshold would provide better predictions.  The answer is no: the in-sample correlation drops off for thresholds above or below 32°F.  It turns out that ice really does melt at 32°F.

Here's this evening's view of the Nenana tripod: the ice is looking dark, so the process is certainly under way.

March Climate Data

I'm not quite back to a normal posting schedule, but here's a quick look back at the March climate anomalies across Alaska.  It was another wetter-than-normal month for the state as a whole, the 5th in a row according to NOAA/NCEI, and indeed only October was drier than normal in the last year.  It was also the 6th consecutive March with above-normal statewide precipitation; the last significantly dry March was way back in 2017.

As in February, the wet weather was focused in western Alaska owing to a Bering Sea trough:

The past three months have certainly been very wet for the Y-K Delta region.  On the flip side, notice how persistent the dryness has been in the southern Panhandle:

 

According to ERA5 data, March was also an exceptionally cloudy month, except in the Panhandle.

Monthly mean temperatures were significantly above normal in the southeastern interior and near the coastline from Anchorage eastward, although not approaching record levels.

ERA5 snowpack data shows a significant excess for most of western and northern Alaska, as well as south-central, but low snow in the southern interior and the Panhandle.

The April 1 NRCS snow survey has some very interesting comments:

"The most exceptional April 1 snowpack in Alaska exists around Valdez. This is a place known for massive snowfall and this year’s snowfall has not disappointed. NRCS Snow Survey uses software to quality control station data based on previous values recorded at a site. The first time our quality control software had to be adjusted for the Upper Tsaina SNOTEL, near Thompson Pass, was in November, when a massive snowstorm eclipsed the amount of snow that had ever been recorded at this site for the date and flagged the data as erroneous. The same thing had to be done in December, and then again in January before we finally set the snow depth higher than the value the station is capable of measuring, which it is currently at. This station was installed in 2002 and has been reading its period-of-record maximum value for most of the year. Right down the road, the Lowe River Snow Course has a much more robust history, and on April 1 it was measured as the highest value in fifty-three years of observation.

Exceptional snowpack continues north from Thompson Pass into the Copper River Basin. Several April 1 measurements in the Copper River lowlands were made as the second highest on record. This basin snowpack has been hearty all year but is not as outstanding as it was last year at this time, when most of the measurements were period-of-record maximums. The other record snow measurements in this report were taken from our partners in Canada, where there are two April 1 records in the upper Porcupine.

The snowpack around Anchorage has been making headlines this year. This highlights a difference in how snow measurements are taken. The Anchorage National Weather Service office at Sand Lake uses a snow board and records the amount of snowfall that falls on the board several times a day throughout the winter. The amount of snow that has been measured is currently the third highest on record and will crown 2024 as the snowiest if a few more inches of snow falls this spring. Snow Survey measures snowpack as an quantity of snow water equivalent (SWE) a site has at a given date. The Kincaid snow course is very close to the Sand Lake office and recorded its sixth highest April 1 reading in its much less robust thirty-three-year period-ofrecord. This is most likely a function of melting during periods of above Normal temperature in February and March.

There are a few places in Alaska with below Normal snowpack on April 1. The measurements taken on islands in Southeast Alaska are below Normal. Several stations are also reporting below Normal snowpack in the interior, where slightly below Normal snowfall combined with warmer than Normal temperatures. In Northwest Alaska Kelly Station SNOTEL has reported below Normal SWE for all of 2024 and is reporting less than half of Normal snowpack on April 1.

In Western Alaska and on the North Slope above Normal precipitation through the winter months hints at above Normal snowpack. Bethel, Aniak and McGrath have had considerably wetter than Normal monthly precipitation totals in February and March. In an effort to understand snowpack in the Kuskokwim basin, the McGrath SNOTEL was installed in 2019. This station is proximal to the no longer measured McGrath Snow Course. The 7.6 inches of SWE reported at the SNOTEL would be considerably higher than the 5.6 inches that is the period-of-record April 1 median for the snow course. Interestingly the reported SWE is the lowest in the McGrath’s five-year history, a testament to how snowy the last five years have been in this region."

Arctic Ice and Temperatures

Last week I mentioned that Bering Sea ice managed to reach near-normal levels this winter despite relatively warm conditions.  The January-March average ice extent was only slightly below the normal of recent decades, continuing the recovery from the very low ice conditions of (especially) 2017-18 and 2018-19.

The Arctic-wide seasonal sea ice maximum was likewise higher than in most recent years, as documented by Rick Thoman on his blog:

https://alaskaclimate.substack.com/p/march-2024-arctic-sea-ice-update

The relatively abundant ice cover is perhaps consistent with the fact that while tropical and mid-latitude temperatures were extremely elevated in recent months, the Arctic did not see quite the same magnitude of warmth that it did during and after the last big El Niño in 2015-16.  Last time around it seemed that the intense El Niño kicked off a several-year period of really excessive warmth in the Arctic, but so far that hasn't happened with the latest (now dissipating) El Niño.

Here's a simple depiction of mid-latitude versus high-latitude temperature trends in December through February, according to the ERA5 reanalysis:

In the mid-latitude sector from 30-60°N, the winter's average temperature was the second highest on record, behind only 2019-20; but for 60-90°N the winter wasn't quite as warm as the three winters from 2015-16 to 2017-18.

Here's a map comparison of this winter (below, top) with the record warmest winters in the mid-latitude (middle) and high-latitude (bottom) sectors.

Clearly the 2017-18 winter had a very strong Arctic focus of unusual warmth, while it was a colder winter in many mid-latitude locations, at least over the continents.  The Arctic was colder in 2019-20 , but there was extreme warmth across Eurasia as a result of a strongly positive Arctic Oscillation (low pressure over the Arctic, and strong circumpolar flow).

In contrast to both of those years, this winter saw very widespread unusual warmth from the tropics to the Arctic.  Below-normal temperatures at high latitudes were largely confined to certain land areas (Greenland, Scandinavia, parts of Russia, and of course some of Alaska), while air temperatures over the Arctic Ocean were significantly above normal.

Given the very modest reduction in the Arctic's warmth this winter, the recovery in sea ice is interesting and, I think, a little surprising.