In this article I decided to include a chart of rolling 365 day mean temperatures from the central England temperature series to illustrate how to use, or maybe abuse, a simple linear trend in a graph of temperatures. The first one is a look at a 365 day moving average of mean temperature for the last twenty years from the 30th of August 1999 (fig 1).
As you can see the 20 year linear trend has flatlined since 1999 with a -0.02°C cooling per decade since 1999. Most people would not believe this is possible when globally the 21st century has already seen nine out of the ten warmest years. All I can suggest is that after the exceptionally warm period that spanned 2006 into 2007 , mean temperatures across central England crashed, with two successive cold winters, which in effect ‘balanced’ out the linear trend.
If you shorten the view of the data by looking at just the last ten years from 2009 instead of 1999 you’ll end up with a graph like this one (fig 2). The linear trend is now not being held back by the two cold winters of 2010 and 2011 but has in fact gained impetus from them with a +1.22°C rise in mean temperature during that period.
The first two graphs illustrate why long-term averages are usually taken over a longer period than ten or twenty years, normally 30 years, like the graph below (fig 3) that covers the last 30 years since 1989.
This probably won’t impress the majority of my readers as the linear trend is quite shallow and shows a warming trend of just +0.09°C per decade, which is much less than the +0.172°C in the estimated global temperatures from the HadCRUT4 series (fig 4). I’ll leave you to figure out the reason why CET hasn’t risen as quickly as global temperatures have in the last 30 years, the reason might lie in the fact that we are an island in the Atlantic ocean.