David Whitehouse: No Underlying Global Warming In Recent Years
In the debate about the significance of the observed global annual average temperature standstill – whose duration now stands at the past 16 years – some have argued that it has little climatic significance. Not only is it shorter than the canonical thirty years used as the minimum to deduce climatic effects, it is also unimportant because the underlying decadal rate of warming is close to the IPCC’s estimate/prediction of 0.2 deg C per decade, and that this rate of warming has remained unchanged over the past thirty years. Thus it is maintained that global warming has not stopped even though there may be a pause in the temperature increase, or as the UK Met Office put it, a recent lower rate of warming. What we have seen in the past 16 years is therefore just variation in the rate of warming and that the underlying rate of global warming is as significant today as it has always been.
The evidence for this is the average global temperature for the past three decades. The UK Met Office in their State of the Climate brochure use an oft-repeated graph that shows this underlying increase in warming.Click on all images to enlarge.
It is obvious there is no pause. It is essentially this data that the head of the UK Met Office, John Hirst, used in a recent lecture. Global warming is taking place at precisely the rate expected.
But there is another way to look at this data.
Firstly, the data in the Met Office graph has been cherry-picked through the process of dividing it into decades. As I have said before the climate knows nothing of a decade which is a time interval that ultimately comes from the number of fingers we have. Imposing arbitrary points on the data set, like taking the start at 1980 and producing data for the 80s, 90s and 00s, is also cherry-picking and has the disadvantage of throwing away the most recent data. There is a different way to look at this data.
Let us look at the last three decades average global temperature but update the Met Office graph to Hadcrut4 (it used Hadcrut3). You will notice it is essentially the same showing a steady increase – the unchanging underlying rate of global warming.
Now, bearing in mind the imposed constraint in the Met Office approach of decades with arbitrary start and end points, take the same Hadcrut4 data but this time use all the available data and work backwards in 5-year integrations.
You will see it tells a very different story. There is now no consistent increase in temperature seen in the data. The mid-point in the Met Office graph is now seen to be the average of the two rather different middle points. What is apparent is that the data show the global temperature has changed between two levels. Each level, of fifteen years, is within one sigma of its mean, and the two levels are two sigma apart. A trendline drawn through the data is clearly unsatisfactory, as this diagram shows.
The data, displayed this way, reveal that far from showing a steady underlying rate of warming the global temperature has had two standstills, with curiously, the 1998 super El Nino delineating them.
These two ways of looking at the Hadcrut4 data (with the cherry-picking considerations) bring out differing aspects but averaging over a decade is giving a misleading impression of what the data actually shows – a well known problem with averaging. The global warming of the past thirty years – the current warm period – does not show an underlying constant rate. Temperature standstills are the norm.
Finally, it has been said that the 16-year standstill observed in the Hadcrut4 data since 1997 has been cherry-picked with its start and end dates. This is not so, the period is simply the answer to the question how far back does one have to go to see significant warming taking the errors into account. In fact, start and end dates are irrelevant, only its duration is important, not where it occurs in the dataset.