Biased BBC Advice Based On Sloppy Statistics

Whatever you think about the BBC’s actual performance in reporting climate change, they are supposed to adhere to the highest standards of impartiality and be able to efficiently gather, assess and represent the state of the science. Only if one starts with a realistic and up to date understanding of the subject can one hope to put into a proper scientific perspective all its developments and weigh the many opinions held about this fascinating and often controversial topic.The BBC should be, or at least aspire to be, the gold standard. So it is depressing to come across such a skimpy analysis, and sloppy use of statistics as in this briefing given to BBC staff by their Environment Correspondent Richard Black. Click on image to enlarge.I will leave Black’s analysis of Climategate, with its several errors in the dates of some of the investigations into it, and the timing of “Glaciergate,” which he says took place before the 2009 Copenhagen meeting, and his crude analogy, and go onto the point in his briefing when he addresses the widely debated topic of the past decade’s pause in the rise of global temperatures.Tomorrow's World“Did it stop in 1998?” Black asks. Is he really unaware of the implications of skewing the data by starting at the warmest year the Earth has experienced in the instrumental period, due to a super El Nino. Most analysts of recent temperature trends would never ask that question. He then goes on to say, “by any common sense definition it ought to be true it stopped in 1997 or 1999.” This is not a logical statement. Even a cursory look at the temperature data shows it is increasing up to 1998, after which there were two cooler (la Nina) years. It is what happened then that the debate is about.Black performs what he describes as a “simple, non-statistical exercise” that first appeared on his blog. He plots decadal trends to show that there has been no reduction in the rate of warming in the past ten years. He takes annual data from NasaGiss and looks at ten-year differences with incremental start points beginning in 1991 showing that only in 1998 – 2008 does it show a negative trend (due to the super El Nino inflating 1998). Note he gets 1999 -2009 increment slightly wrong.As Black admits it is a simple test, but he clearly thinks it is appropriate to show such an analysis as part of his briefing to a room of BBC editors, producers and journalists. The problem with it is that it makes the rudimentary mistake of ignoring the short-term variations and noise in the data resulting in spurious trend estimates that, as statistics often does in the wrong hands, obscures more than it illuminates. A more scientific and statistically preferable approach is to start in 1991, using monthly data, and plot ten-year regression lines. It is obvious that they are converging on zero for the past decade – the exact opposite of what Black told his audience.  Whatever it means, and whatever its cause, the pause in global warming is a real effect. Black says that variability in the annual data means one probably shouldn’t do such an analysis. I concur.A Kick Up The EightiesAfter this amateurish display things get a little more confused. When describing the data (HadCRUT3v Global data this time) Black spoke of a “relative plateauing” in the past decade, even though his crude trend analysis given a moment before didn’t show it. He then said, “you could make a case that global warming has plateaued, but if you are going to say that you would also have to say global warming has plateaued there, and there and there.”He was pointing at the much shorter standstills seen in the data in previous decades. These are well understood, and not comparable to the past decade. In two cases they are due to volcanic eruptions (Mt Pinatubo in 1991 is obvious in the data, and there have been no such eruptions in the past decade). It is highly misleading to compare apples and oranges in this way. Science can explain the slight pauses seen in the two decades before this one, though it has a harder task explaining the 1940-1980 standstill.The point is that previous flat periods, the cause of which is debatable, occurred before the date given by the IPCC at which mankind’s influence on the global climate was dominant (sometime around 1960 – 80). A hiatus in warming nowadays is a somewhat more important part of understanding what mankind’s influence on our planet is, hence the current considerable discussion about possible decadal influences on climate.I'm Sorry I Haven't A ClueWhat Mr Black with his “non-statistical exercise” did not do is what one would have expected a BBC correspondent to do. That is, reflect the scientific literature concerning the temperature pause of the past ten years. The are many, many examples, and it is not now widely contested in scientific circles. Only a few months ago the Proceedings of the National Academy of Sciences published a peer-reviewed article that began: Data for global surface temperature indicate little warming between 1998 and 2008. Robert K. Kaufmann, at al., “Reconciling anthropogenic climate change with observed temperature 1998–2008.” PNAS, June 2, 2011.The pause has been discussed in Nature Climate Change, Science, and acknowledged by the Royal Society and the UK Met Office, here and here. Even Mr Black himself has previously written about the causes of the past decade’s hiatus.Black also shows a graph he used in a 2007 article, ‘No Sun Link’ to climate change. The sense of triumphalism in this article, as well as its inadequacies, I have gone into before.The point is that even when it was fresh and not four years old, the graph of cosmic ray intensity and of rising temperature was out of date. Can it really have escaped Mr Black the considerable debate, the uncertainties and new assessment about the sun’s influence that has been taking place following the Sun’s very unusual behaviour in the years after he wrote his 2007 article.My experience is that BBC Editors are as intelligent and as fast-thinking an audience as you could get anywhere. Quick to pounce on strained logic and inconsistency, especially in a news report. That is why they are usually the gold standard. But they are not scientists.This is a dismaying standard of scientific literacy from a BBC correspondent. Following Black’s presentation the BBC audience went away with the opposite impression of what is the case. Given the severe cutbacks the BBC is experiencing at the moment it would be like saying there will be more jobs, not fewer. I do hope that when those cuts are explained to the staff that a somewhat more sophisticated use of statistics is used.Feedback: david.whitehouse@netzerowatch.com