Sunday, October 13, 2013

As Close to "Normal" as It Gets

One of my pet peeves it the use of the word "normal" to describe climate averages.  Even The Weather Channel has raised concerns about this, yet it is commonly done by the National Weather Service and television broadcasters to describe the weather statistics on any given day.

It is perfectly "normal" in the midlatitudes for the weather to fluctuate.  The average weather for any given day does not equal normal.

Over the past 31 days, we have had a very good example of this.  In fact, during the past 31 days we have had about as close to "normal" weather as we can have in late sumer and early fall.  The temperatures at the Salt Lake City International Airport have fluctuated up and down with a gradual cooling trend as we moved through September and early October.  There have been days with above average temperatures, and days with below average temperatures, but the fluctuations have not been extreme (only one record was tied during this period).

Source: NWS
In the end, after all these ups and downs, the average temperature for the 31 day period is within 1ºF of the 30-year average.  You don't get much closer to average than that.  The airport also received 1.02 inches of precipitation, which is just a shade below the 30-year average for the period.  

So, this has been a remarkably average fall, with relatively "normal" weather.  The end result is that the Wasatch Mountains seem to look as they should for mid October.  There is a skiff of snow in the high country (see above) and the leaves are past peak at upper elevations and near peak at mid elevations.

This aspen grove (~7800 ft) is still a bit green and just approaching peak 
Near 8500 ft we're probably near or just past peak
This view up Butler Fork shows that the highlands between Big Cottonwood and Mill Creek Canyons are
now past peak, but leaves hang on in some areas.
There are some exceptions to this rule, which may be partly related to aspect and wind exposure.  One thing is for sure, this weekend is probably the last gasp of decent leaf peeping above 7000 feet.  By next weekend, most of the aspen leaves will be gone.

Finally, I have to comment that we are in the early stages of what is expected to be a very dramatic increase in global and regional temperatures.  While the past 31 days is within 1ºF of the average for the past 30 years, that probably puts it about 1.5ºF above average for the 20th century (I could run exact numbers, but hey, it's the weekend).  Thus, average for us is not average for our grandparents.  


  1. Thanks for a good post. One of the things that has always bothered me about our use of "averages" is that, as you note, the "average" temperature listed on the TV and in newspaper is the 30-year average. Obviously this custom skews recognition of climate change. Is this 30-year norm worldwide practice? Has there ever been a discussion among meteorological professionals that you all should lobby the media to, say, provide both the 30-year average (calling it that by name) but also, say, the 20th-century average? Also, when did 30 years become the definition of "average"? Thanks!

    1. Ultimately, it is simply a protocol recommended by the World Meteorological Organization. This is probably more than you want to know:

      My guess is that most of the time you these days you are seeing the 1981-2010 average. The NWS would do well to include such information in anything that includes an average statistic so that we know for sure.


    2. I was originally going to comment on this topic before I got to the bottom, so I will tag on.

      I learned within the past year that technically "normals" and "averages" are not synonymous, despite some people using them interchangeably, and/or assuming that a normal is the same thing as a 30-year average.

      A "normal" is a statistic created by the National Climatic Data Center. They are listed in the daily NWS climate reports and are thus widely used. Yes, they are based on the most recent 3 decades of data, but it is not a simple average. The data (especially temperatures) go through an adjustment process, of which I don't know the details, just that it is supposed to make the data more "representative." Sometimes the differences can be more significant than one might expect.

      Throw in the possibility of using a period of record average (which can be easily accomplished if one has access to the ACIS interface), and you create quite a challenge when someone asks "What is the average number of 90 degree days per year in X City?" It could be 13 (NCDC "normal"), 15 (pure 30-year average), or 20 (period-of-record average). :-\

      Also I'd speculate that the 30 year number allows one to compare different station's normals in an apples to apples context, in addition to being the magical statistical (n=30) threshold. It wasn't too long ago that many stations didn't have much more than 30 years of reliable data as well.