Social Channels


Additional Options

Date Range

Receive a Daily or Weekly summary of the most important articles direct to your inbox, just enter your email below:

Roz Pidcock

Roz Pidcock

23.10.2012 | 4:40pm
ScienceThe Mail on Sunday, David Rose and climate misinterpretation: Round two
SCIENCE | October 23. 2012. 16:40
The Mail on Sunday, David Rose and climate misinterpretation: Round two

Last week was a big week for skeptics and climate scientists debating the long-standing (but wrong) claim that global warming has stopped. Just to recap, skeptic journalist David Rose provoked a huge response following his article in the Mail on Sunday last week, entitled “global warming stopped 16 years ago – and here’s the graph to prove it”. He suggested that the alleged finding was contained in a Met Office report that had been released under the radar. 

Although the Met Office and a host of scientists, commentators and bloggers criticised Rose’s analysis, he published another article in this week’s Mail on Sunday sticking to his original story. Rather than treading more carefully around the scientific data this time, Rose made no acknowledgement of his previous mistakes and even threw a few more erroneous claims in for good measure.

The new article includes a climate Q and A and an invitation for readers to “decide what the real facts are”. So we do just that by taking a look at what he says this time round.

The “report” and the ‘official’ Mail on Sunday graph

In the first article, Rose implied that new global temperature data came from a report released by the Met Office “without fanfare or publicity” – a claim which The Met Office rejected on its blog. It said:

“…the Met Office has not issued a report on this issue. We can only assume the article is referring to the completion of work to update the HadCRUT4 global temperature dataset compiled by ourselves and the University of East Anglia’s Climate Research Unit.”

As Rose clarifies in his new article, the original graph that he showed of surface temperature data did not come from the Met Office. In fact, as was obvious, he made it himself, clarifying that the Met Office “put out the figures from which we drew our graph ten days ago”. Even so, he still calls his graph “a new official world temperature graph” – a strange turn of phrase, given that it is clearly a Mail on Sunday graphic.

Daily Mail _recap
The graph in question.

Even though neither the graph nor the interpretation of the data came from a report released by Met Office, media outlets all over the world unquestioningly picked up Rose’s version of events under headlines like ‘ Global Climate Warming Stopped 15 Years Ago, UK Met Office Admits‘.

There are some issues with the graph. The first data point is the temperature anomaly for September 1997 (the difference from the 1961-1990 average temperature) but it seems to be labelled as the start of the year. This makes each of Rose’s data points appear six months earlier than it really did and also means that the time period he refers to is 15 years, not 16 years. Climate data analyst and blogger Tamino and the Grantham Institute’s Bob Ward have both written about discrepancies between the actual Met Office data and Rose’s version.

Significant warming

Data inconsistencies aside, the fundamental argument that Rose returns to, and one that crops up time and time again on the skeptic blogs, is that global warming has stopped. The specific argument that Rose repeats in his new article is:

“[S]ince January 1997 there has been no statistically significant warming of the earth’s surface”.

The Met Office has stressed – including, apparently, to Rose – that records of 15 years or so have limited relevance for analysing global climate change. Short-term excursions from the long term temperature trend are not unusual in the climate record, and so do not mean that the warming trend apparent over the last century has stopped. We wrote more about this here.

The term “statistically significant” also has a specific mathematical meaning when it comes to detecting trends in data. Whether a trend is statistically significant can vary depending on the time period it is measured over. And a bit more data either side can change the outcome. But Rose doesn’t explain this to his readers – instead, he opts for sweeping statements.

“Championship cherry-picking”

Rose’s choice of starting date is central to his argument. Rose claims in his new article that he picked 1997 as a start date because that was when the period of warming ended. He says:

“Some critics claim this newspaper misled readers by choosing start and end dates that hide the continued warming. In fact, we looked at the period since 1997 because that’s when the previous warming trend stopped”.

Either Rose didn’t actually test whether the previous period of warming changed at all in 1997, or thought that no one else would. Climate data blogger Tamino – who specialises in this kind of analysis – examined the Hadcrut4 data and wrote on Sunday:

“It turns out that the trend using just the data since mid-1997 does not disagree (statistically) with the trend prior to mid-1997.”

Tamino shows the graph below. The 1997-2012 period that Rose uses in his argument is marked in red. The previous warming trend that Rose refers to it marked in blue. If both time series are treated as whole, the same trend continues as before.

Source: Tamino

Having identified 1997 as his starting year, Rose even chooses to start his analysis in September of that year rather than January. In September 1997, a warm phase of the natural ocean and atmospheric cycle known as the El NiÅ?o/La NiÅ?a oscillation (ENSO) had started to kick in – which hadn’t happened by January. This marked the beginning of a short period of unusually high global surface temperatures, usually lasting a few months to a year.

We are now in a fairly neutral year in the ENSO cycle, which means neither unusually warm (El NiÅ?o) or cold (La NiÅ?a) conditions. Beginning Rose’s time series with unusually warm temperatures – meaning that later temperatures are being compared against a high starting point – inevitably skews the trend to show a slower rate of warming between the start and end points. Tamino argues that Rose’s choice of dates is:

“not just cherry-picking, it’s championship cherry-picking.”

Ocean warming

Rose claims in yesterday’s article that “while CO2 levels have continued to rise since 1997, warming has paused”. As we noted in our blog post last Monday, this ignores warming in other parts of the climate system, notably the oceans. But in the new article, Rose appears to dismiss the possibility that ocean heat uptake is occurring altogether, saying:

“…other scientists say that heat has somehow been absorbed by the waters deep in the oceans. However, the evidence for this is contested and there are no historical records with which to compare recent deepwater readings”.

Here Rose appears to be dismissing thousands of temperature measurements that form the World Ocean database, which has found that the ocean (from the surface to 2000 metres) accounts for approximately 93% of the warming of the earth system that has occurred since 1955. The image below shows that the amount of heat going into the upper ocean (0 to 700 metres), the deep ocean (700 to 2000 metres) plus the land/ice/atmosphere have all risen steadily between 1997 and 2012.

Nuccitelli _OHC_Data
Source: Nuccitelli et al. (2012)

Natural factors

In response to the part of the Q and A that asks “When will warming start again?”, Rose answers:

“It is likely that in the 2020s, the Atlantic cycle – currently in warm mode – will also flip to cold, so that for some years both the Pacific and Atlantic cycles will be cold at the same time. When this happens, temperatures may decline like they did in the Forties.”

We assume that here Rose is referring to the natural climate cycles the Atlantic Multidecadal Oscillation (AMO) and the Pacific Decadal Oscillation (PDO). The AMO is currently in a warm phase and scientists are uncertain when it will end – traditionally each phase of the cycle lasts around 35 years. The current warm phase started around 1990, so the best guess is that it could end in the 2020s. The PDO typically remains in each phase for around 20 years, and has over the last few years began a cool phase.

As far as we can see, there is a lack of peer-reviewed literature to suggest what the effects of the two cycles coinciding might be, especially when superimposed onto the recent temperature increase due to anthropogenic greenhouse gases.

Medieval warming

In the section of yesterday’s article entitled ‘But isn’t the world still much warmer than at any time in recorded history?”, Rose claims:

“Ever since it was published on the cover of the IPCC third assessment report in 2001, the ‘hockey stick’ graph showing stable or declining temperatures since the year 1000, followed by a steep rise in the 20th century, has been controversial.”

Rose claims that the graph in question fails to include a period in earth’s history known as the Medieval Warming Period (MWP). We wrote more about this subject here, but to summarise, the available data suggests that warming during the Medieval Warm Period was not as globally widespread as the warming seen since the 1970s, and that average global temperature during the period probably did not match the warming of recent decades.

2013 and beyond

Next year, the Met Office will again update global temperature data. If the update shows that the 1997-2012 warming trend is statistically significant, we wonder if Rose will give that news the Mail on Sunday treatment – or whether he’ll allow the data to be released, in his own words, “quietly on the internet, without media fanfare”.



Expert analysis directly to your inbox.

Get a Daily or Weekly round-up of all the important articles and papers selected by Carbon Brief by email.


Expert analysis directly to your inbox.

Get a Daily or Weekly round-up of all the important articles and papers selected by Carbon Brief by email.