Saturday, October 20, 2012

Putting Down the Winter Snowpack

The Fairbanks area received three to five inches of snow over the course of several days early this week, with the result that everywhere in the area now has a solid snow cover. In spite of the mild weather early in the month, this is right on schedule: the long term median date for establishment of the permanent winter snow cover is October 18th. It is possible, but high unlikely, that the current snow cover could be reduced to a trace. Since 1929, twice has there been two inches of snow on the ground (the current snow depth at the Airport) on October 20th, only to have it melt back to a trace later in the month. Snow cover is defined here to be a snow depth on one inch or more at the standard time of day for snow depth reporting (traces do not count) and each day thereafter also has one inch or more on the ground through March 31st.

A couple of things stand out on the chart. First, in the past twenty years there has been much less variability in when the winter snow pack was established compared to previous decades (only one year more than 10 days from the median). Also note that it's a lot more common to have very late establishment of snow pack than very early. Only two autumns has the snow pack been established more than 15 days early (1956 and 1992), but seven times has the snow cover been more than 15 days late (though none since 1982). This includes the remarkable autumn of 1934, when the thin snow cover of November was obliterated by the super Chinook of December 3-8, 1934 and then there was no measurable precipitation until after the day after Christmas, resulting in the only "brown Christmas" in Fairbanks weather history.


  1. I am curious as to why the weather forecasters at NOAA are predicting a warmer-than-average Nov-Feb winter for northern Alaska. I understand that temperatures are currently much warmer than average there, but this is primarily due to the lack of sea ice. Once sea ice regains its full coverage, shouldn't temperatures return back to normal? Thanks

  2. That is a really good question Trung. Right now the lack of sea ice (compared to pre-2002) is a major factor in coastal North Slope temperatures. It is, if you will, seriously "stacking the deck" toward above normal temperatures due to all the latent heat release. Once the sea ice is back in place, then that source of heat is reduced. It could still be warmer than normal, but the deck isn't "stacked". Make sense?

    As for why the Climate Prediction Center is tilting the odds toward above normal temperature forecast for the whole winter, I can't say for sure. However, one tool CPC uses is the recent past. If, say, 10 of the past 15 winters have been well above normal, then you might tilt the odds toward well above normal.


  3. Rick, Trung: One other thing the CPC might be looking at is the CFS computer model forecast which shows much-above normal temperatures persisting almost everywhere over the Arctic ocean for Dec-Feb. Here's a link:

    It doesn't show the warmth extending south of the North Slope, though.

    I wonder if the ocean water under the ice can still provide significant warmth in the first part of the winter because the ice is thin, and presumably this year it will remain thin for longer than usual.


  4. Richard,

    I'm sure CPC is taking into account the CFS forecast. CFS version 2 is well worth looking at and has done a good job over the past year (for Alaska; I can't speak to elsewhere). Now why the CFS is forecasting that is a different question.

    There is some heat flux through the ice but more from cracks and leads. Sea ice is not like ice on a pond, all smooth and even. The fact that it's now all first year ice makes it that much easier to fracture. You can see this regularly on infrared satellite images in mid-winter. Over the ice and right on the coast is often much warmer than just a few miles inland.

    Thanks for the comment,