Hiding the Decline: Climategate Demystified

by Brian Dunning
December 12, 2017

Nearly everyone's heard about Climategate, a scandal said to prove that climate scientists deceptively changed temperature data as part of a scheme to hoax the world into thinking global warming is real. If you have heard of Climategate, you've probably heard that the scientists involved were all investigated but cleared of any wrongdoing, thus suggesting that the investigating agencies were complicit in the hoax. You may have even heard the key phrase from one of the emails: "Hide the decline," an apparent revelation that the scientists knew temperatures were actually declining and had to do something to hide the fact. But how many of us know the actual details of what happened? Fortunately, the Climategate scandal is not all that complicated, and an insider's view is the subject of today's show.

It all happened in November 2009. The backup email server at the Climatic Research Unit of England's University of East Anglia was hacked, through methods which have not been made public for security reasons. A 160MB file of about 1,000 emails and their attachments appeared on a server in Russia. The first known parties to download the archive were all editors at climate change denial websites such as WattsUpWithThat. Headlines quickly followed. From Fox News:

Why You Should Be Hot and Bothered About 'Climate-gate'

From The Telegraph:

Climate change: this is the worst scientific scandal of our generation
and
Climategate: the final nail in the coffin of 'Anthropogenic Global Warming'?

From the Houston Chronicle:

Climategate: You should be steamed

And so on, and so on. And yet, when it seemed so clear from these headlines that global warming had been completely obliterated and revealed to be a hoax, the editors of the world's most prestigious science journal Nature quickly wrote their own response. It said in part:

The e-mail archives stolen last month from the Climatic Research Unit at the University of East Anglia (UEA), UK, have been greeted by the climate-change-denialist fringe as a propaganda windfall... proof that mainstream climate researchers have systematically conspired to suppress evidence contradicting their doctrine that humans are warming the globe. This paranoid interpretation would be laughable were it not for the fact that obstructionist politicians in the US Senate will probably use it next year as an excuse to stiffen their opposition to the country's much needed climate bill.

And that is, of course, exactly what happened — exactly according to the hopes of the hacker and the bloggers the files were sent to.

Here's what happened. Cutting straight to the chase, 99% of Climategate focused on one single email. It had been sent 10 years earlier, in November 1999, by Phil Jones, head of the Climatic Research Unit, to five colleagues. It is six sentences long, but this one sentence is the culprit:

I've just completed Mike's Nature trick of adding in the real temps to each series for the last 20 years (ie from 1981 onwards) and from 1961 for Keith's to hide the decline.

This has been interpreted as "I used a trick to hide declining temperatures." If true, it would indeed be damning, not just to Phil Jones, but to climate science as a whole, given the who's-who of climatology who were the email's recipients. But is that a truthful analysis of the email? Does it really prove that climatologists knew that the Earth is cooling, and so invented tricks to deceptively portray the data to advance their warming agenda?

To find out, let's parse out this sentence that rocked the world. The email was discussing the author's progress creating a temperature graph intended for the cover of a report made for the World Meteorological Organization. The graph covered the past 1000 years, overlaying three data sets showing the fluctuating global temperature anomaly within a range of about 1 degree. And, of course, it jumps up sharply in the latest 50 years, about 3/10 of a degree.

Although we have many lines of evidence for historical temperatures which all track well with one another, most aren't complete. Most are available over a given time range. For example, an ice core tells us about the time period for which that ice has existed. Cave formation data covers the time period for which that cave was wet. Tree rings go back as far as the trees were alive. Instrumental data goes back as far as thermometers have been reliable. There is plenty of overlap, which allows us to say that these lines track one another, but in order to build a complete timeline, we rely on multiple data sets.

"Mike's Nature trick" refers to what climatologist Michael Mann did for a 1998 article in the journal Nature. In the article, Mann and his co-authors displayed a reconstructed historical data set, known as MBH98. This temperature data extended only up until 1980. Before the article was published, Nature's peer reviewers suggested also displaying the modern instrumental temperature record, which extended all the way until the present, for context. This was done. The two curves were shown on the same graph, each clearly labeled, and the data for both curves was already public domain. This, and this alone, constituted the entirety of "Mike's Nature trick."

So by employing Mike's Nature trick in his graph for the World Meteorological Organization, Phil Jones was simply adding the instrumental data. That's all.

This brings us to "hide the decline". What was Phil talking about here? Turns out there was indeed a problem with one of the data sets he wanted to include, and it's called the divergence problem.

To understand the divergence problem, we first have to make a few basic points about dendroclimatology — the science of using tree rings to indicate historical temperatures. Dendroclimatology is one of the many lines of evidence for determining the planet's climate history prior to the age of thermometers; others being glacial ice, corals, cave formations, sea levels, glacial extent, and others. Generally a tree ring represents one year, and when we look at the tree rings of a given year from all over the planet, we can see how climate varied across different regions. Each tree ring gives us two primary pieces of data: its width and its density. Generally, a wider ring usually means a wetter growing season. A longer, warmer summer results in a denser late seasonal growth period, manifested as the dark part of the ring. So the density of the ring tells us about temperature and cloud cover. Historically, all of these lines of evidence have tracked one another very well. Since the dawn of reliable direct measurements with thermometers in the 1800s, both tree ring density and tree ring width have tracked extremely well with temperatures all over the world.

And this leads us to the divergence problem. In the 1990s, it was noticed that after 1960, the relationship between temperature and tree rings started separating. While temperatures rose, tree rings continued to indicate cooler temperatures, in defiance of all other measurements. We don't see this divergence in any other lines of evidence; tree rings are the only climate indicator that has done this, and only in the northern hemisphere. Why is it happening? Well, that's why we call it the divergence problem. We don't know. Likely candidates include fractionally diminished sunlight caused by sulfate aerosol pollutants, warming-induced drought, increased ultraviolet radiation caused by ozone depletion, even loss of permafrost. But the bottom line is that in the past few decades — and only in the past few decades — tree rings in the northern hemisphere have indicated cooler temperatures than what the thermometers have shown, so we know them to be wrong. If you want to combine all the lines of evidence to build the best possible picture of climate history, you have to stop using dendroclimatology right around 1960 for temperature, or else your graph will be wrong and misleading.

This is what Phil Jones was referring to when he wrote "from 1961 for Keith's to hide the decline." He was talking about Keith Briffa's tree ring data, known to be inaccurate since 1961. Jones elected to display the correct instrumental temperature data instead. And if you look at some global warming temperature graphs that overlay multiple lines of evidence, you'll see that the tree ring temp series stops around 1960, just as the modern instrumental series stops before 1860 or so. To make a useful graph, we show the data that we have.

Deniers often claim that we only display tree ring data until 1960 because we know it proves temperatures are actually declining since then. That's because they're unaware of the divergence problem, or they think it's something we made up as an excuse for the Climategate revelations. This claim can be easily disproven by picking up the literature. The first two major papers about the divergence problem were both published in 1995, four years before Phil Jones wrote the infamous email, and fourteen years before the alleged scandal. Proof that dendroclimatology data became unreliable after 1960 was fully public and well known to everyone in the field, and had been for years.

Claims that "hide the decline" meant "hide the fact that global temperatures have been declining" are also unraveled by the timeline. Phil's email was 1999, and 1998 had been the hottest year on record, peaking a global rise throughout the 1990s that nobody disputed, as it was all instrumental data. There had been no decline of global temperatures to hide.

If you do any search for this online, you're quickly going to see that most of the climate change denialist writing contracts Jones' email into the following: "Mike's trick to hide the decline." As is obvious by now, Michael Mann's multiple data sets on one graph has nothing at all to do with the tree ring divergence problem. That's why neither Phil Jones nor anyone else who knew what they were talking about ever used the phrase "Mike's trick to hide the decline"; there was no such thing. The phrase had to be manufactured by deniers by cherrypicking words out of unrelated parts of Phil's email.

So what was the fallout of Climategate? Well, as a propaganda stunt to discredit climate science, it was all too effective, with many people still believing even today that it did indeed reveal global warming to be a hoax. Eight different committees investigated the scientists involved and none found any evidence of bad science, misconduct, or anything else that touched upon the science of global warming. However, the emails selected by the hackers were often those that caught the scientists in embarrassing moments. Years of battling harassment by climate deniers had led the scientists to sometimes become defensive, and the hacked emails revealed moments when the scientists stonewalled requests for data from certain groups; and in one case, recommended deletion of certain emails should future requests be made for them. In the words of one committee, they showed a "consistent pattern of failing to display the proper degree of openness". Understandable given the constant harassment, but not excusable. Also, not relevant to the science, which despite some public distrust, survived Climategate unscathed.

So the next time you hear a friend explain that global warming was proven to be a hoax by the Climategate emails, you now have the tools to break down why they think that's the case. You're probably not going to change their mind — most of us cling to our politics like religion — but you may be able to get them to reexamine their sources.

By Brian Dunning
Follow @BrianDunning