NOAA have recently been promoting that November 2012 was the 333rd month in a row with above average global temperatures, and this has been widely picked up by the media (e.g. here). But, how useful is this statistic?
Firstly, this statistic uses the global mean temperature from one dataset (GHCN v3.2.0), and assumes that “average” is the whole of the 20th century. The answers would vary slightly when using a different dataset, but the choice of average period is key. The 20th century is an arbitrary choice (as is any such choice) – a smaller number of months would be found if a more recent averaging period such as 1961-1990 or 1986-2005 was used instead. So, the number of months depends entirely on an arbitrary choice, which ensures the statistic is far from robust.
Secondly, but far more importantly in my view, this global statistic does not match how people experience climate, which is on the local scale, mainly by looking out of the window. Recent work suggests that “people can perceive and adapt to aspects of climate variability and change based on personal observations“. My personal experience matches this.
Below, the same NOAA data and averaging period is used, but now presented regionally, and shows how many months since a below average month in each available grid cell, i.e. locally. The picture is very different with the majority of regions having a cooler than average month within the past 2 years, and for many regions it is less than 3 months. Interestingly there are slightly larger numbers in the tropics, perhaps because the variability tends to be smaller there.
This type of global statistic is damaging, precisely because it does not tally with how people experience climate change and variability. Much better to take the opportunity to educate and increase understanding about how the climate varies from year to year.
[Many thanks to Doug McNeall, John Kennedy & Scott Mandia for useful discussions.]