Thursday, January 19, 2017

Cold Snap

Clearing skies allowed temperatures to drop to very low levels in most parts of the interior in the past 48 hours, with Fairbanks airport reaching -51°F yesterday morning.  The high temperature for the midnight-to-midnight period was only -41°F.  These are the coldest conditions in Fairbanks since late January 2012.  Many other locations, distributed very widely across the interior, dropped well into the -50s, with the "winner" being the Kanuti Lake SCAN site at -61°F as of 5am today.

The following map, courtesy of NOAA/NWS, shows reported low temperatures below -50°F in the 24 hours ending around 5am today.

As we noted before, the deep trough over Alaska has imported very cold air aloft, so the low surface temperatures are entirely to be expected under clear skies with very little solar insolation and light winds.  The Fairbanks soundings have measured 850mb temperatures around -20°F in the past couple of days, and it's not unusual to see an inversion of 30°F or more from the surface to 850mb at this time of year.  Under the right circumstances the inversion can reach 40°F or more, but that's more common under high pressure.  The sea-level pressure (MSLP) has actually been considerably below normal in the past several days.

The relationship between MSLP and temperature inversions is illustrated by the pair of charts below, which show the 850mb and surface temperatures on high-MSLP days (top chart) and low-MSLP days (bottom chart) according to the winter-season Fairbanks balloon soundings since 1981.  There is considerably more scatter in the relationship when MSLP is high, because the surface conditions are more decoupled from the free atmosphere, and accordingly strong inversions are more common; notice that there are quite a number of days in the top chart with surface temperature more than 40°F colder than 850mb temperature.  In contrast, this almost never happens when MSLP is low, and it's not coincidental that surface temperatures almost never get below -40°F when MSLP is low.  Interestingly, yesterday had one of the lowest MSLP readings on record for a -40°F surface temperature in the Fairbanks sounding, and this attests to the strength of the upper-level trough.

Here's the 500mb analysis from yesterday morning, courtesy of Environment Canada:

Finally, here are some interesting FAA webcam photos of shallow ice fog layers yesterday at Chalkyitsik, Arctic Village, and Livengood.



  1. I wonder if the duration of cold snaps has been abbreviated since...? Seems that may be the case as they used to linger longer. May have been already answered here.


    1. Great question, Gary. I don't recall looking at this specifically before. I'll write more later, but yes there is evidence that the coldest spells have become shorter in recent decades.

      Here's one symptom of this: from 1930-1960, one in 5 winters saw a daily mean temperature below -50°F; and there were 35 such days in total. From 1981-2015, one in 5 winters saw a daily mean temperature of -45°F; but there were only 11 such days in total. It appears that the modern climate produces fewer days at the bottom end of the temperature distribution even after accounting for warming.

    2. Gary, I did a write-up in 2015 on the length of warm/cold spells in Anchorage at 850 mb. What I found was that cold spells were marginally shorter, but warm spells were notable longer.

    3. Thanks for the fill and link Brian. I enjoy your Blog and thanks for taking the time to inform the rest of us.

      Richard I guess the reason I (and Brian) ask about duration (maybe it's been different for large zones of Alaska?) is the implied evolution to the Arctic via changes to our climate and perhaps Polar Amplification.

      Perhaps the effects of zonal and meridional flow on temperature spells have changed if the above theories are as correct as they appear to be.