Sunday, April 26, 2020

Length of Summer - Brian B Blog

Most readers will be familiar with Dr Brian Brettschneider of UAF, who has become a sought-after authority on Alaska climate.  Brian recently wrote an article on his own blog that I thought would be worth linking here:

Are Summer/Winter Longer/Shorter Than They Used To Be?

Using a simple method, Brian demonstrates that the "summer" and "winter" temperatures of a prior climate period (1960-1989) in Alaska have become considerably longer and shorter respectively in the most recent 30-year period (1990-2019).  The changes have been particularly dramatic in winter across Alaska, and most notably of course on the North Slope.

Brian's work provides a nice visualization and easy-to-understand interpretation of temperature changes over the past several decades.  In my opinion, it would also be interesting to extend the work to build in the different perceived length of summer/winter at different latitudes.  For example, the introduction acknowledges that Houston has a longer summer than Omaha, and Chicago has a longer winter than Oklahoma.  Most people would say Alaska has a very long winter and a short summer; and despite the dramatic warming on the North Slope, winter certainly hasn't disappeared there, as Brian's simple method would imply.

Obviously it would be difficult to come up with a good metric to capture subjective opinions of summer and winter, but perhaps a threshold temperature could be added that would demand a "summer" or "winter" classification, regardless of how long that season might be.  For example, perhaps:

- Normal daily mean temperature above 27°C -> summer
- Normal daily mean temperature below -10°C -> winter

The winter threshold would then produce a (1981-2010 climate) winter season from October 28 to March 21 in Fairbanks, and October 22 through May 3 in Utqiaġvik.  Perhaps readers would like to comment on this or other ideas to refine the analysis without losing too much of the attractive simplicity.


  1. It's subjective and likely based upon personal experience. For me anything that averages daily temps over 0*C in Fairbanks is either spring or fall. Anything over +10*C is summer. Grim but true.


  2. Grim indeed. So that leaves about one month each for spring and autumn. Interesting perspective, thanks.

  3. Yes that's about right in my experience...a month for each maybe and often + or -. When I use the term average that may be gracious. Staying above freezing daily closely guarantees winter conditions like snow and ice are at risk.

    By ~April 20th it should be spring enough to support greenup and fall can end soon after September 20-25th. Varies by year. Photoperiod/solar angle plays a factor in fall more than Spring. Even with 12+ hrs in later March there's usually a delay in warmup. Hit that equinox in September and prepare yourself for winter.

    Leaves turn yellow in mid to later August and new buds can form in later April if that's any indication of seasonality.


  4. Thank you for sharing my post. Defining the seasons is a rather intractable problem. What Kansans consider winter would not seem very wintry here in Alaska. Vice versa for summer conditions. Defining a set length for a reference period and measuring how long those prior conditions are experienced today felt like a good place to start, but is just that, a start. It breaks down though for tropical latitudes, where winter and summer have little meaning with respect to temperatures.

    I do like the idea of perceived season length, but then we get into the issue of defining some initial conditions. If we say summer has a daily mean of 27°C, then no station in Alaska has ever had a single summer day. :)

    1. Yes, intractable and perhaps futile in the end. Of course I didn't mean to suggest that a fixed temperature threshold could serve by itself, but it could provide a useful extension to the 90 days in places with a long winter/summer. The problem is that perceptions differ so wildly that no simple scheme will seem reasonable everywhere.

      But I agree - your analysis was a good place to start and the results are informative, well done.

  5. I really like what Bousted et al. did with the winter severity index (AWSSI). They defined winter using four criteria that are applicable everywhere - perhaps more applicable for mid latitudes, but satisfactory for high latitudes. A companion analysis for summer would be great.

  6. Meanwhile, we in the Northern Hemisphere await the Final Warming often promised to those that outlast winter.