There now seems to be a great deal of confusion about what is happening to global temperature. Some of the confusion is the result of sheer ignorance of statistical methodology on the part of scientists as well as the media and the general public. This is not so much a lack of knowledge as the adherence to incorrect ideas about statistics. Sometimes organizations which should be deseminating information are instead are deseminating misinformation. And sometimes the confusion stems simply from a reluctance to look at the data.

Here is the record of annual average global temperatures expressed as deviations from the long term average. These data are provided by the National Oceanic and Atmospheric Administration (NOAA). The other sources of such data have similar but not exactly the same figures. First here is a graph of the data.

Since the precise values are relevant, here is the tabulation of the data.

The AGT reached a peak in 1998, probably because of an El Niño condition. That peak was surpassed in 2005, but by only 0.0183 of 1°C. On a typical household thermometer in which one inch represents 10°C that difference is about the diameter of a human hair. That would not be of much significance except 2005 can be labeled as a year of recond high temperature.

Here is what NOAA reported.

The 2005 global temperature was statistically indistinguishable from the standing record set in 1998. One data set, in use at NCDC since the late 1990s, produced a global annual temperature for 2005 that was slightly below 1998. An improved data set, which incorporates innovative algorithms that better account for factors such as changes in spatial coverage and evolving observing methods, results in 2005 being slightly warmer than 1998.

On January 14th 2011 NOAA released the average global temperature figure of 0.62 °C above the 20th century average. It was given only to two decimal places. The more recent updates of the historical data seem not to be available, but there is a more than good chance that the figure for 2010 tied the 2005 figure only because they were rounded up to two 0.62 °C. Given NOAA's rediculous record mongering if the 2010 figure had exceeded the 2005 figure by any minuscle amount it would have been reported as an all time record temperature. Given that it was not possible to report a record high the next best thing was to manufacture a tie with the past record by artful rounding.

Government agencies and the popular media have perpetuated the notion that a new record temperature is evidence of an inexorable trend. This is not true. A statistical variable that has no trend but random variations about a constant level will over time achieve new record levels. This is just an example of a larger sample having a greater sample maximum. The AGT over the period 1998 to 2010 appears to have no upward trend. It might even have a slight downward trend.

If one wants to see a variable with an upward trend consider the data in the increase in the concentration of carbon dioxide in the atmospher over the 1997 level shown below.

The data for this graph also came from NOAA.

To drive the point home consider the two graphs above superimposed.

Perhaps the past twelve years is too short of a period infer anything about long term term trends. If one looks at the long term record, as shown below, there is clearly evidence of a cycle.

The pattern is one of a cycle imposed upon a long term trend. The long term trend is 0.5°C per century; a definite trend but not one of catastrophic warming. It is probably due to human activities but not entirely or even predominantly due to carbon dioxide. For more on this analysis see Cycles and Trends in Average Global Temperature.