Cold At Last

Cold air aloft and clear skies allowed temperatures to drop to the coldest of the season so far across much of western Alaska in the past couple of days, and -40° was breached for the first time at a measuring location in the state.  This is much later than usual.  Several sites dropped below -35°F earlier in the month, including -39°F at the Ruby 44 ESE co-op site, but yesterday's cold was more significant.  Here are some 24-hour low temperatures ending at midnight last night.

The coldest reported (not on the map) was -49°F at the Kaiyuh RAWS near Kaltag.  McGrath reached -45°F, the coldest since the pre-Christmas cold spell of last year.

But more unusually, Bethel reached -30°F, the coldest since February 2017.

In other news, Anchorage snowfall continues to track at record levels, with 79" so far this season.  This is above the normal total for the entire winter.

On a different note, while fiddling with climate data over the break, I took a look at the relationship between temperatures aloft and temperatures at the surface in the depths of winter, for all the major balloon sounding sites in Alaska.  For temperatures "aloft" I used the 850mb pressure level, which averages about 4000-4500 feet above sea level at this time of year.

There's a robust positive correlation, of course, between 850mb and surface temperatures at all locations, but it does vary quite a bit across the state.  The site with the highest correlation of average December-January temperatures is Nome:

The lowest correlation is at Utqiaġvik, but this is at least partly because of the differing trends at the surface and aloft: surface temperatures have warmed more rapidly because of sea ice loss.

In general we expect lower correlations for more interior locations that have strong and near-permanent surface-based temperature inversions at this time of year, and higher correlations at maritime sites where inversions are less common.  This is broadly what we find:

The highly maritime climates of Kodiak, Cold Bay, and St Paul Island are at the top right, with Annette Island also quite similar.  On the far left, with strong inversions, are Fairbanks and McGrath, and these sites have a weaker relationship between surface and 850mb temperatures.  Anchorage is right in the middle.

The only one that surprises me is Yakutat, with a weak correlation despite a rather maritime climate and positive lapse rate (i.e. no temperature inversion on average).  More investigation would be needed to illustrate why this is the case.