Location: Round Rock, Texas, United States

I maintain: a good sense of humor; a belief in hard work; and eclectic tastes in food, art, and politics.

Thursday, January 13, 2005

Seasonal variability and mean temperatures

It seems that a change in the weather is coming tonight, and apparently none too soon for those who love cold weather. Even those who do not like cold weather are ready for colder temperatures because they believe that the winter will be worse unless the early winter is cold enough, whatever that might be. Even meteorologists fall into this way of thinking about some facets of the weather that might be "normal seasonal variability." Also, today's news on one of the networks originated from Minnesota, and of course the story included the obligatory view of those residing in Minnesota that it needs to be cold to give us a change to suffer appropriately. While colder (and adverse) weather means money for some people and businesses, the usual result of colder weather is that it becomes the topic of conversations. Which then reduces its importance in inverse relationship to the number of times a particular aspect of the weather is discussed in groups of two or more people. Probably most of us have our own preferences and pet peeves about the weather, but mine is somewhat easier to pinpoint. For me, as for those in Minnesota (I suspect), enduring the harshest weather is bearable as long as I can receive credit for having survived something. What causes the weather to seem unbearable is when someone, usually a local weather person who seems to be especially untrained in statistics, starts off by use of the term "the high for today should be," as if there is some kind of imperative involved. What seems to happen is that these people can never understand the simple difference between averages (means) for a day, means for the high, and means for the low temperatures. For example, if the temperature is 50 degrees F., and the average mean is 40 degrees F. For the day, you can count on the fact that some weather person will say that the high should only be 40 but at 50 we are above what we "should be." Actually, if the high mean is 50, and low mean is 30, and the daily mean is 40, we might be at the exact mean temperature that could be expected. But expected in statistical terms is not the same as "should be," because expected can fail to occur because of the daily variability that is always associated with the means that are computed and maintained by the weather service firms, government or private. While this may seem to be picking on a straw man, it actually reflects a general belief that something called "reversion to the mean" will always take place. While it frequently does happen that over time the averages seems to be achieved many times, there is only the theory that says the averages will revert to the mean by either having lower minimums or higher maximums, or whatever type of variability is required to reclaim the mean. So the next time you hear the local weather person, who seems to have some simple comments to explain a common phenomena called the weather, remember you are probably listening to some minor facts laced with numerous theories regarding those facts. Why weather persons might as well be academics with their heads in the clouds, or was that the ivory tower? That's 30.


Blogger jim said...

This topic has been brewing for years and well stated.

12:21 AM  

Post a Comment

<< Home