Winter Sunshine

Noontime sun and 13 below at Central. Photo courtesy of the FAA.

Dramatic Temperature Swings

Interesting temperature variation the past day in Fairbanks-land. Here is a plot of hourly temperatures from a number of locations.

The max temperatures early in the morning make a useful case study: Eielson reach their max temperature when the southeast winds broke through ahead of the trough, so are hour or so earlier than the other valley locations (in spite of Eielson being the eastern-most of these stations), where the max temp occurred with a short pulse of weak southwest wind following the trough. At elevation (Keystone Ridge), there was no spike in temperature at all, where most of the wind occurred Tuesday evening.

High temps (not just hourly) occurred 5-7am:

Eielson AFB: 26F
Keystone Ridge: 21F
Fairbanks Airport: 16F
Woodsmoke PWS: 11F (near North Pole)
Goldstream Creek: 7F

Solar Heating or Lack Thereof

Here's a plot of the daily duration of solar heating at 65N (latitude of Fairbanks), defined as the duration the sun is 5° or more above the horizon. Note the plot extends for 15 months so to show the winter minimum. We are a week into the two months with no significant solar heating. Changes in surfaces temperatures during this time are modulated by horizontal movement of air (advection), vertical mixing (mostly wind) and presence of absence of clouds significant to reduce the continuous cooling of the top of the snow pack (long wave radiation).

Big warm-up on the Way

Trung asked about the big warm-up in the current forecast in a couple days. Here's why:

Big pattern change is underway for Alaska and will bring much milder weather to the Interior by midweek. First, here is the 500mb heights and 850mb temperatures for this past Saturday morning (from the ECMWF). This shows a closed low west of Anchorage and a fairly chilly airmass over most of mainland Alaska; the blue colors are 850mb (~4500 MSL) temperatures of -16C or lower.



Now, here is the forecast from the 12Z Sunday ECMWF valid at 3am AST Wednesday Nov 30th. Deep southwest flow from the north Pacific covers most of the state, with the chilly air having retreated the Brooks Range. This could wind up being a fairly snowy pattern for Fairbanks-land, but as always in these situations, if the winds aloft are just a little bit too southerly, then we will "chinook out". At this point though it looks to me like the best forecast is for some snow the middle and end of this week.

Inversion

The Sunday afternoon sounding from Fairbanks shows a decent 15C inversion in the lowest 500 meters. This is reflected in the late afternoon temperatures, which remain in the 20s below in the valleys and a little above zero in the hills. As usual, most of the increase is in the lowest 100 meters.

Big Inversion

A pretty decent inversion has developed overnight, with valley temperatures pushing 30 below and higher elevations near zero.

10AM Sunday temperatures include:

Woodsmoke PWS: -33F (near North Pole)
Eielson AFB: -29F
Fairbanks Airport: -27F
Goldstream Creek: -27F


Cleary Summit: +2F
Little Chena Ridge +1F (south of the Steese Highway)
Gilmore Trail PWS: -1F (1440' MSL)
Keystone Ridge: -2F

Cold in Context

Here's another view of the recent cold weather: the daily mean temperature plotted against the 1981-2010 normals; grey shading of plus/minus one standard deviation.



Here are standardized daily departures for 2011. Looked at this way, the recent cold snap was about as far from normal as the hot weather in late May. The coldest day, the 17th, with an average temperature of -35F, was 2.7 standard deviations below normal, almost the same as warmest day in, when the 28th was 2.9 standard deviations above normal. The height of cold snap was a couple days longer than the hot spell in May, but the total "excessive" (the area under the curve) will likely be somewhat greater in May than in the cold snap.

Thanksgiving Day

305pm November 24th from UAF West Ridge

Cold Persists near the Alaska Range

Unlike the Fairbanks area, clear skies have prevailed near the Alaska Range for past 36 hour. Here's the 1pm photo from Denali National Park right near the train station. The lack of clouds have allowed the cold temperatures to persist. At the Denali Park AWOS (partially visible in the photo just beyond the brown building), the high today was 20 below and low 36 below. At the DVCA2 RAWS, which is near the Wilderness Access Center and at a slightly lower in elevation than the AWOS, the high was -27F and the low -38F.

Recap of the November Cold Snap

The record setting cold snap of November 2011 is history. So, without the media hoopla, here’s what happened. It's important to emphasize that the "news" here is the early occurrence of the cold snap, rather than the absolute values of the temperatures. The exact same temperatures anytime between Thanksgiving and Valentines Day would be unexceptional.

The mean 500mb heights (30dm contours) for November 15-21 (courtesy of ESRL and NCEP/NCAR Reanalysis) show a classic pattern for cold weather in Interior Alaska, with a high over Chukotka and fairly low height northerly flow across most of mainland Alaska (I’ve added centers and arrows to indicate wind direction.



In the Fairbanks area, this was easily the coldest pre-Thanksgiving week of record. At the Airport, the lowest 7-day average temperature was -30.1F for November 15-21. The previous record was -26.5F was set November 7-13, 1989. For November as a whole, this was the third coldest event; November 24-30, 1909 and the same week in 1927 were colder.

Record lows were set at the Airport on six days (November 16th did not have a new record) and record low maximum temperatures were set November 16-19. The high temperature of -30F on the 17th is the second earliest high of -30F or lower in the winter (November 9, 1989 is the earliest). The low temperature of -41F on the 17th is the fifth earliest occurrence of -40F in the winter. Here's the plot of November record lows:




Elsewhere around the area, temperatures varied pretty much as is typical for this time of year. The average temperature for November 15-21 at the North Pole cooperative station was -37.7F, with six days with lows in the -40s and an absolute low of -49F on the 17th. At the other extreme, Keystone Ridge cooperative station, at 1600’ MSL, had an average temperature of -13.4F the same week and an absolute low of -28F on the 21st.

Most places with a long period of observations set six or seven daily records. Lowest temperatures for the cold snap include:

University Experiment Station: -41F
Eielson AFB: -41F
UAF West Ridge: -40F

As more cooperative info becomes available I'll update the extremes.

Impact of Clouds on Temperatures

The last 24 hours have featured a textbook example of the impact of clouds on valley temperatures in Interior Alaska in winter. Here is a plot of the hourly temperature at the Fairbanks Airport. Variations in temperature are completely related to changes in clouds cover: clears out for a few hours, the temperature drops 20 degrees. Clouds return, and the temperature pops right back up. In contrast, here on Keystone Ridge, the temperature has varied by 5 degrees in the same time, -1F to -5F

Inversion Collapses: Cold Snap Ends

Temperatures aloft warmed dramatically over the past 36 hours, and persistent cloud cover and a moderate pressure gradient have taken their toll on the inversion. Here is a plot of the lowest 1000 meters from the morning sounding from today and Monday.


At 2am today (Nov 23) the temperature at the Fairbanks Airport rose to 2 above. This effectively ends the cold snap. The numerical models have not distinguished themselves in glory with the end of this event, and that is directly related to the inability of the models to effectively resolve the processes in the very near surface (boundary layer). That said, the airmass remains chilly and if and when clouds break temperatures will again fall in the valley. However, the unprecedented nature of this cold snap is over. I'll have a wrap-up in the next couple days.

Fairbanks Cold Snap Update

For the week ending Nov 21, the average temperature at Fairbanks was -30.1F. This is, by far, the coldest 7-day average temperature so early in the season and the coldest week in November in the WB/NWS era (since 1930). In the early years, when observations were done by cooperative observers, the week ending on Nov 30th in 1909 and 1927 were colder.

Clouds and some light snow have arrived and overnight temperatures have been mostly in the 20s below in the valleys. On exposed hillsides and channeled valleys a breeze picked up too. Here at Keystone ridge at 6am, -10F with northeast winds about 15mph with gusts to about 25 mph. Air quality is good!

Solar Noon in Barrow

Here's the FAA webcam shot looking south from the Barrow airport. The photo was taken at 116pm AST, which is very close to solar transit.

Updated Records

Here's am update showing the record lows that have been set this month at Fairbanks. The low of -38F this morning breaks the previous record of -35 set in 1904.

Dramatic Cooling

Here are plots of the temperatures in the lowest 1000 meters from the upper air sounding at Fairbanks from 3am AST Sunday and 3am AST Monday. While the surface temperature was only about 3C cooler Monday morning that Sunday morning, temperatures aloft fell much more: over 10C at 100 meters above the ground; the temperatures aloft are now as cold or colder than last week.

The NOAA-19 Polar orbiter infrared image from 5am AST Monday shows lots of cloud scraps across the interior but not much organized yet.

Coldest Early Season Week of Record at Fairbanks

The airmass cooled dramatically during the day Sunday but few clouds showed up, allowing for another very cold day. At the Fairbanks Airport, the morning low of 34 below, but the temperature fell to 37 below during evening, breaking the previous record of 36 below set in 1993.

For the week of Nov 14-20, the average temperature at the Fairbanks Airport was -27.6F. This is the lowest weekly average temperature of record so early in the season. The previous record was -26.5F set Nov 7-13, 1989. It is also very close to the coldest week anytime in November since 1930, when the Weather Bureau began observations: currently that record is -27.9F set November 21-27, 1948. That record will likely be broken after today. Prior to 1930, there are a couple of end-of-November cold snaps that far exceed what we've had thus far: In 1927, the average temperature Nov 24-30 was -36.6F and the same week in 1909 had an average temperature of -34.3F

Cold Moving East

The eastern Interior hung on to clouds a couple days longer that central and western sections, but skies have cleared now and temperatures fallen. The low of 46 below at Northway ties the record low for November 20th. Other low temperatures include 46 below at the Tok PWS, 45 below at Chicken and 43 below at Eagle before the wind picked up.

This morning's photo is the 11am view from Lake Minchumina, looking southeast toward Denali and the sunshine.

How Unusual is this Cold Snap?

An interesting, and increasingly asked question is: "how unusual is this cold snap?"

First some more numbers, just looking at Fairbanks (1904-2010):

Saturday (Nov 19) was the fifth day (in a row) with the daily minimum temperature 35 below or lower. Only four other Novembers have had five or more days with low temperatures of 35 below or lower.

Thus far, three days have had maximum temperatures of -20F or lower (Saturday will likely be the fourth). Eleven other Novembers have had three or more such days and six have had four or more.

The lowest temperature (thus far) has been -41F; this been exceeded in nine other Novembers.

However, this event is occurring not toward end of November, but around the middle of the month.

At the close of Saturday, the 7-day average temperature will be about -22F; there have been five other years that had a 7-day average temperature of -20F or lower ending November 20th or earlier. The coldest 7-day average temperature ending on or before November 20th is -26.5F, which occurred on the week ending November 13, 1989. In second place is Nov 11-17, 1956 at -23.9F.

Another way to look at this, though probably of dubious value, is to ask when the last time a cold snap set four daily record lows? Answer: September 1992, a "black swan" event if ever there was one in Interior Alaska. I say dubious because three of the four records broken were what are sometimes called "low hanging fruit", as can be seen in this plot of November daily record lows for Fairbanks:



Disclaimer: While I have doubts about the veracity of some of the pre-1930 data, it's what the NWS uses for daily records, so I include it here.

The three record lows of 33 below were the only records that high after November 4th. From this point onward only one day has a record lower warmer than -36F.

So, is this event unprecedented; no. Unusual? You bet; similar cold snaps at this time of year have occurred only a handful times in the past 106 years, and this may approach the coldest early season prolonged cold snap (before late November), though I'd guess at this point the 7-day record low mean of -26.5F set in 1989 will stand.

Strong Inversion

Back home now, so more useful posts, and I can comment too!!

Here's a plot of the lowest 500 meters of the upper air sounding from Fairbanks this (Saturday) morning. A fairly typical "strong inversion" situation, worth 30+ degrees F. Readily reflected in the surface observations.

At 8am…

Fairbanks Airport -35F
Eielson AFB -37F
Woodsmoke PWS -41F (near North Pole)

Cleary Summit -3F
Keystone Ridge -5F
Parks Highway RWIS -8F (at FNSB border)

Looks like this will change early next week as a tight cold low drops south from the high Arctic into the western Interior. I expect the inversion will largely break with clouds and light snow. There is a good chance that the hills will have somewhat lower temperatures than earlier this week while the lowlands will be substantially warmer than it has been.