The Mail On Sunday And The Met Office
In response to an article in the Mail On Sunday that points out the absence of a recent temperature rise in the Met Office’s newly released Hadcrut4 global temperature database the UK Met Office released a statement that is misleading.
The Mail On Sunday article uses the Met Office’s Hadcrut4 database that was updated from 2010 to the present day last week.
We live in the warmest decade of the instrumental era (post-1850), and most of the warmest years have occurred in the past decade, but what the Met Office ignores to say is that, at present, we live on a temperature plateau – there is no recent upward trend in global temperature.
The Met Office says that the world has warmed by 0.03 deg C per decade since 1997 based on their calculation of the gradient in the Hadcrut4 dataset. But what the Met Office doesn’t say is that this is statistically insignificant. The gradient of the trendline in Hadcrut4 is very sensitive to the start and end dates used as temperatures vary significantly month-to-month, so the Met Office is being misleading in quoting trendlines for a particular start and end date without taking into account how the scatter of the data, the errors in the temperature measurements, and short-term changes affect the statistical confidence in the resulting trendline.
Trendlines from 1997 to August 2012 vary between 0.04 to 0.02 deg C per decade with an associated error of 0.04 deg C per decade. This has to be considered along with the error in annual global temperature measurements of 0.1 deg C. Hence there is no case to be made for a statistically significant increase in global temperatures as given in the Hadcrut4 dataset between 1997 and August 2012.
Quoting trendlines without errors can mislead. For instance the trendline between January 2002 and August 2012 in Hadcrut4 is negative, being minus 0.04 deg C per decade: Between January 2003 and August 2012 it is minus 0.05 deg C per decade – that is global cooling. Would the Met Office be happy to quote such figures in the same way they have for 1997 onwards and state that the world has cooled in the past decade? Only when the errors are incorporated, which the Met Office did not do, can these be seen to be statistically insignificant.
The Met Office also says that if they were to calculate a linear trend from 1998 (a strong El Nino year) to August 2012 it would show a warming more substantial that 0.03 deg C per decade. Actually the warming since 1998 is the same – 0.03 deg C per decade – and again statistically insignificant.
The year 1997 – roughly the start of the recent temperature standstill – is not cherrypicked. Before that year there is a statistically significant increase to 2012, after 1997 there is not.
The Met Office says the 15-year standstill is not unusual. This is true but again the Met Office is being economical with the truth. The IPCC concluded that the period 1960-80 marked the start of mankind’s domination of the Earth’s climate via greenhouse gas forcing. The period before 1960-80 the IPCC regarded as being solely due to natural factors. In the pre 1960-80 period there was a standstill between 1940-80. In the post 1960-80 period there was warming between 1980 – 96 and a standstill thereafter. The mankind-dominated era has only one standstill, which is becoming the dominant global climatic feature of this era.
Only a few years ago the Met Office said that temperature standstills of a decade were common (about one in eight decades), but that temperature standstills of 15 years were not supported by their climate models. They appear to have altered their view as the observed temperature standstill lengthens. The Met Office’s track record in predicting global temperature changes has been dismal.
The Met Office says that climate change can only be detected in multi-decadal timescales. In the three decades since the IPCC said that mankind dominated the Earth’s climate there has been equal timespans of warming and temperature standstills. Which one do they consider to be more significant?
It is disappointing, if not misleading, that when the Hadcrut4 data was announced in March, with data only available to 2010 (a warm El Nino year), the Met Office promoted it with a press release and briefings to journalists. They told Louise Grey of the Daily Telegraph that the Hadcrut4 data showed that the world had warmed even more than expected in the past ten years and that the warming between 1998 – 2010 was 0.1 deg C.
When the full dataset was available, in the past week, showing global temperatures to August 2012, and telling a very different story, no press release was produced.
In an official response to comments on the Met Office website criticising the Mail on Sunday article by David Rose the Met Office have said regarding their trend of 0.05 deg C in Hadcrut4, “…the figure is not intended to show significant warming has occurred. We agree with Mr Rose that there has only been a very small amount of warming in the 21st Century.”
This is clearly in contradiction to the original message of the Met Office’s release, i.e. that there had been warming in the past 16 years. One could ask what is their evidence that the warming in the 21st Century is “very small” when by their own admission their figures are not intended to show statistically significant warming had occurred. As we have pointed out if one uses the Met Office technique of deducing warming one finds that the trend in the 21st Century is negative – i.e. there has been cooling.