Hottest July on Record - Really!

There has been a great deal of "noise" made concerning the average July temperature for 2012 which broke the old record set in 1936.  My first thought about these kinds of statements is that if you have a list of 100 numbers, one of them has to be largest among the 100.  The fact that the highest number in the list occurred in the most recent timeframe, does NOT indicate any statistical significance!  Therefore, I wanted to study these data to understand if there is any statistical significance to July 2012 temperatures.

I went to the NOAA.gov site to retrive the July Average Temperatures by year since 1895.  Indeed, July 2012 was in fact 77.56 Degrees F and outside the Upper Control Limit.  The previous record was in 1936 at 77.43 degrees F which was just inside the UCL.  Here is a X, MR Chart for the entire data set.





Notice the yellow highlighted vertical line which indicates the beginning of a 14 year period where 13 of the 14 years are all above the overall average of 74.402.  This indicates a short term shift in the average.  Beginning 1944, the results return to vary around the overall average and remain consistent at this level until 1998 when another run of 12 out of 15 years again occurs.  Within this statistically different period again the new record year occurs.

This then begs the question: "are these two 14 year periods statistically different from each other which might indicate a warming rise over these 82 years.

In the chart above, each of the two "higher" periods each has its own average and limits.  The first thing to notice is that the 1936 and 2012 temperatures are not outside the Upper Limit and, therefore, not unique.  The more important conclusion is to notice that these two averages are NOT STATISTICALLY DIFFERENT.  This is indicated in the lower left corner of the chart which also states that the year to year variation is not different between the two periods.  In spite of the fact the most recent 15 year period temperature is 0.3 degrees higher than the earlier 14 year period, these two averages are NOT statistically different.  However, rest assured that the media and even some scientists would suggest that this 0.3 degree increase is an indication of "global warming".  The "records" in 1936 or 2012 are NOT UNIQUE.  What needs to be explained is the years 1930 and 1998 when these 14 years periods of higher temperatures began.  Trying to explain what happened in each of the these two 14 year periods would be useful but difficult to connect to a phenomenea of "gradual warming".