Dear Editor,
The EPA is speaking nonsense when they talk about “global temperature.” It makes no more sense to calculate an average temperature for a whole planet than it does to calculate the average telephone numbers in a phone book. Temperature, like viscosity and density, and of course phone numbers, is not something that can be meaningfully averaged.
If enough accurate surface temperature measurements existed to ensure reasonable planetary coverage (they don’t), it is however theoretically possible to get some sort of temperature statistic (though interpreting its significance would be a challenge).
But what averaging rule would you use to handle the data from the roughly 6,000 temperature sensing station that are currently used? Mean, mode, median, root mean square, what? Science does not tell us. In fact, for some groups of close temperature measures, one method of calculating an average can lead to a determination of warming while another can lead to a conclusion of cooling.
