Friday, November 30, 2012

Where is the Coldest Place in Alaska?

Regular reader Richard commented that Chicken frequently is cited as a place reporting very low minimum temperatures and asks what is the coldest place in Alaska. This is a question I regularly get, so here you go.

The first thing to keep in mind is the lack of good climate data over vast swaths of Interior and northern Alaska. Since the mid 1950s there have been only two Weather Service Offices in Interior Alaska (Fairbanks and McGrath), an area of something like 300,000 square miles. This is reasonably supplemented by several long term FAA stations: Northway, Big Delta/Fort Greely, Tanana, Bettles. Barrow is the ONLY long term climate site still operating on the North Slope, an area of about 85,000 square miles. There are a handful of long term sites (from a variety of observers) that are useful: Galena and Eagle, and some in the immediate Fairbanks. Everything else is hit or miss. Places you might think should have good long term climate data simply do not (e.g. Fort Yukon, Manley Hot Springs).

The lowest official temperature of record in Alaska is -80F, set in January 1971 at Prospect Creek, which was a pipeline construction camp at Mile 135 Dalton Highway (yes, there are many claims of lower temperatures). The lowest monthly average temperature (NCDC 1981-2010 normals) is Umiat, on the banks of the Colville River, about 140 miles southwest of Prudhoe Bay, with a January normal temperature of -21.3F, and a winter (Dec-Feb) normal of -19.3F.

Umiat, which started out as a Navy station during World War 2, is these days a logistics supply base for oil exploration and general access to the central inland North Slope. One family called Umiat home from the mid 1970s til about 2000, but before and after it is similar to Prudhoe Bay, in the it's no one's home but there are people there working in one capacity or another.

Umiat has the lowest winter mean temperature of any "settlement" in Alaska. That I am confident of (I wrote a paper about this back in the mid 1980s). However, based on 25 years of looking at weather satellite imagery of Alaska, I'm confident there are places upriver from Umiat that are colder. In fact, Umiat is often on the northern edge of the cold pool of air that develops on the north side of the Brooks Range. My guess is that some places southwest of Umiat (maybe near the confluence of the Colville and Killil Rivers) probably average 3-5 degrees cooler than Umiat. All these areas though are cold because it is hard to "warm up" rather than the super low temperatures that have historically been measured in the Interior. The lowest reliable temperature at Umiat, and the lowest I know of on the North Slope, is 66 below in February 1987. Lots of places in the Interior have been colder.

Turns out that today is a good day to discuss this topic. To the left is a NOAA-16 Polar Orbiter infrared image from 630am Friday. There is not much wind and hardly any clouds over the Interior, so the cold air pooled in valleys shows up nicely. To start with, there is nothing especially cold about Prospect Creek, or any of the Brooks Range valleys (including current or former climate stations at Chandalar Lake, Wiseman, Wild Lake, Coldfoot, Arctic Village). They all can get very cold in the right circumstances. Prospect Creek holds the record low simply because there happened to be a climate station, which was open for less than ten years, during a deep cold snap.

The Yukon Flats can be very cold. However, it's at low elevation (generally below 500; MSL), and areas north of the Yukon River can be breezy at times, breaking up the surface temperature inversion. The Yukon River Valley above the Flats can also be very cold, though there is practically no human habitation. The only community is near the Alaska-Yukon border at Eagle, which is at 900' MSL but is prone to downriver winds, breaking the inversion. The areas downriver from Eagle are less prone to wind. but are at slightly lower elevations. 

The Upper Tanana Valley also gets very cold: it is about 1000 feet higher elevation than the Yukon Flats, and there are areas that do not get much drainage wind (e.g. Tok area, Northway and the Tetlin flats in general). However, the Upper Tanana Valley is comparatively close to the Gulf of Alaska (so gets clouds from that direction), and is also far enough south (Northway is at, gasp, 63N) that the sun has some impact on daytime temperatures except for very close to the winter solstice.

Which brings me to my candidate for the coldest area in Interior Alaska: the valleys of the Fortymile County. The Fortymile County is basically a high elevated plateau with winding rivers deeply cut into the plateau. The general elevation of the plateau is up around 2000-2500' MSL, with valleys, often very narrow, several hundred feet lower. The area was a focus for gold mining a century ago, but today is very sparsely populated. The Taylor Highway (BLM brochure map on the right), which is not maintained in winter, cuts through the Fortymile County, connecting the town of Eagle, on the Yukon River to the Alaska Highway. Chicken is the only community in the Fortymile County to have a post office.


The community of Chicken, and the current cooperative observer are in the valley of Chicken Creek, a small tributary of the Mosquito Fork Forty Mile River. Chicken does have the lowest mean January (and winter) average temperature in Interior Alaska, with a 1981-2010 normal January temperature of -20.9F (normal Dec-Feb temperature -16.3F).  There is nothing unique about the location of the coop station, Having been there, it looks to me to be completely representative of the area. It gets so cold because it is a high elevation valley well protected from any wind most of the time, maximizing radiational cooling. However, in spite of current NCDC normals, there is in fact less than 20 years of usable data from Chicken. If any currently active climate station can break the Prospect Creek record low, it's Chicken (which has been to 70 below twice in the past decade).

Chicken though is almost certainly not the coldest place in the Fortymile County. The short-lived O'Brien Creek co-op station, about 40 miles north of Chicken just north of the Fortymile River bridge was generally a bit colder (higher elevation, deeper valley) than Chicken. Because it was in a deep valley, it goes several months without actually seeing the sun.  It is in these kind valleys the absolute lowest temperatures are most likely to occur.

The lack of human habitation and the difficulty of getting automated equipment to work and record accurate extremely low temperatures makes documenting the extreme temperatures of this region difficult.




13 comments:

  1. Fascinating, Rick - thanks so much. Very nice discussion.

    I have noticed extremely cold readings on occasion from Galbraith Lake, nestled against the north slope of the Brooks Range - this supports your contention about colder temperatures south of Umiat.

    Do the infrared satellite temperature estimates have sufficient spatial resolution (and precision/accuracy) to differentiate between temperatures in some of the coldest spots? I wonder if one could possibly use IR data from many clear nights to identify candidate locations for the true "pole of cold". Just a thought.

    ReplyDelete
  2. Richard, Thanks.

    Yes, 1km resolution Polar Orbiter imagery is useful for this kind of work and helps inform my opinions. In January 1993 we used calibrated POES resolution to identify sub-80 below temperatures in some valleys southeast of Arctic Village.

    Rick

    ReplyDelete
    Replies
    1. Interesting. Can you recommend a good site for viewing the Polar Orbiter imagery? I would like to become more familiar with it and look for some of these cold spots.

      Delete
    2. Richard,

      For Alaska-centric polar orbiter imagery, the NWS Alaska Region website has the best real-time imagery:

      http://www.arh.noaa.gov/poes.php

      For all kinds of high latitude imagery, the Geographic Information Network of Alaska (GINA) at UAF is a great resource:

      http://sv.gina.alaska.edu/index.pl?embedded=1

      Hope this helps,

      Rick

      Delete
    3. Thanks - great links, very helpful.

      Delete
  3. Thanks for the informative article. I'm wondering if the area south of the Brooks range close to the Canadian border could possibly be the coldest spot in Alaska. Old Crow, Yukon is consistently one of the coldest spot in Canada in the winter and it is rather close to the Alaskan border. Would it make sense for a location within its proximity on the Alaskan side to be just as cold. What do you think?

    ReplyDelete
  4. Trung,

    Old Crow and the Crow Flats are indeed very cold. However, the flats end just east of the Alaska-Yukon border, and the Porcupine River valley is fairly narrow where it crosses into Alaska. Based on satellite imagery it gets downriver winds on a regular basis.

    Rick

    ReplyDelete
  5. I live just a bit outside of Tok, which is currently in the grip of a cold snap (people have gotten readings down to -60 recently, with rumors of sub -70 temps around the corner). I think your theory of the isolated deep valleys in the Fortymile is a good one, but as far as populated areas go, I'd say the large region between Tanacross and Beaver Creek, YT seems to be the most severe. I've often thought that the depths of the Tetlin refuge must be capable of some very extreme temps, but no one is there to read them! In any case, this whole area (Fortymile, Upper Tanana, White River valley) is sort of like the North American equivalent of the infamous "pole of cold" in the Sakha Republic of Siberia (Ojmjakon, Verkhoyansk, etc.). In any case, after many years of living in Fairbanks, I'd say that the Tok area definitely has a more brutal winter climate.

    ReplyDelete
  6. Robichaud74 — You're certainly correct that Tok-land is the coldest population center in Alaska. The greater Upper Tanana Valley is one of the North American cold poles. The Yukon Flats is another, the Kanuti Flats in the Koyukuk drainage is yet another (e.g. Allakaket is significantly colder than Bettles). An uninhabited mini cold pole is certainly the Mosquito Flats west of the Taylor Highway. Follow the clear sky infrared images for a few winters and these recurrent cold areas pop-out. Thanks for your comments.

    Rick

    ReplyDelete
  7. So is it plausible that an extreme subartic climate (Dfd, Dwd),the type that's found in the Sakha Republic, can also exist in Alaska somewhere in the deep valley pockets of the Fortymile Country. A deep valley that never sees the sun sounds like a place that can sustain -40F temps even through warming periods.

    ReplyDelete
  8. Trung, the deep valleys of the Fortymile County stay cold for a long time, but ultimately the area is much closer to a source of clouds and moisture (with resulting surface warming) than anywhere in Siberia. Check out the data from Chicken the past 10 years to get an idea of how cold it can stay.

    ReplyDelete
  9. What's interesting though is that a place like Ojmjakon is actually a bit closer to the Sea of Okhotsk than places like Eagle, Dawson or Mayo are from any moderating body of water, and the mountain ranges separating it are also significantly lower. So, I think the extremes observed there have more to do with the power of the Siberian high interacting with the particular topography than the distance to moderating influences. The truly deep interior of Siberia (the vast stretches of taiga between the Lena river in western Sakha and the Putoran plateau in Krasnoyarsk Krai) while quite cold, doesn't seem to suffer from quite such extremes. In any case, I think some of the past readings from Sakha are highly questionable, at best. One other area I do wonder about is that basically uninhabited region around the Yukon/NWT border, due east of Mayo. There seems to be some pretty low land there surrounded by fairly rugged mountains, with lots of canyon/valley topography as well, and it's about as far away from moderating influences as you can get in North America. Sorry if I'm a bit off-topic here!

    ReplyDelete
    Replies
    1. Oymyakon sits on a high plateau (>2200' MSL) and Sea of Okhotsk partially freezes up during winter. Several of the NWS folks in Fairbanks have visited there. The town is at a lower elevation that the airport/met station, and of course gets colder than the official obs site.

      No doubt many of the old Siberian records are questionable. I seem to recall an 80 below or so at Oymyakon at some point in the past few years.

      Rick

      Delete