āGlobal warming on pauseā.
IF YOU have been reading the papers of late, you may be under the impression that global warming isnāt proceeding as expected.
While most mainstream media have been careful to point out that the apparent lack of recent warming is probably just a temporary hiatus, a few outlets have suggested there is more to it than that. āThe climate may be heating up less in response to greenhouse-gas emissions than was once thought,ā .
Advertisement
What is going on? Has global warming really slowed or stopped in recent years? If so, why? And does this mean the world wonāt warm as much as previously expected?
Claims that global warming has stopped are nothing new. The vast majority donāt stand up to scrutiny, but the latest talk appears to be different. This time itās climate scientists themselves talking of a slowdown ā and in the past few months they have even begun to publish papers about it.
This past decade has been the hottest since records began ā thatās not in dispute. However, the average surface temperature of the planet seems to have increased far more slowly over this period than it did over the previous decades. The rate of warming was just 0.04 Ā°C per decade from 1998 to 2012, significantly lower than the average 0.11 Ā°C warming per decade since 1951 (see āHow much has warming slowed?ā).
Yet this doesnāt necessarily mean that climate change has stopped, any more than meant that it had accelerated. Instead, a standard explanation is trotted out to explain these changes in pace: natural variability.
Surface temperatures go up and down all the time because of changing winds and currents, and phenomena such as volcanic eruptions. This variability can easily obscure the underlying warming trend. Remove the known contributions of natural variability from the observed surface temperature, and you see a much steadier warming trend (see āPutting the brakes onā).
Yet glibly blaming natural variability for the recent surface slowdown is unlikely to satisfy those who are genuinely sceptical about global warming. How, they might ask, can we be sure that the plateau in surface temperatures is due to natural variability masking the underlying warming trend, and not because warming has actually slowed or stopped?
To answer this, it helps to think about heat energy rather than temperature. The reason the planet has been warming over the last century is because rising levels of greenhouse gases act like extra blankets, reducing heat loss from the top of the atmosphere.
In terms of heat, there are three possible reasons why Earthās surface temperature hasnāt increased much recently. The first is that the sun has been getting dimmer, meaning less heat is arriving. The sunās heat output rises and falls in an 11-year cycle and by spacecraft such as SOHO show it did dip particularly low recently.
The second reason could be that more heat than usual has been escaping from the top of the atmosphere. One possible cause is increased levels of sulphur aerosols in the atmosphere. These aerosols donāt prevent the sunās rays entering our atmosphere, but they do reflect more of the sunās heat back into space. Sulphur aerosols are produced by volcanic eruptions ā one of the big causes of ā as well as from coal burning and other human activities. Sure enough, , mainly due to lots of small volcanic eruptions.
Lastly, it is possible that the planet has still been gaining heat, but that more of it has ended up somewhere other than the lower atmosphere, whose temperature we focus on. The most obvious culprit is the ocean. Water covers more than 70 per cent of the planet and the stuff has a huge capacity to absorb heat: around 3000 times as much energy is needed to warm a given volume of water by 1 Ā°C as is needed to warm the same volume of air.
Observations show that a of the heat energy gained by the planet since 1971 has ended up in the oceans, with another 4 per cent absorbed by land and ice (see āWhere is the heat going?ā). So all the surface warming since 1971 is due to just 2 per cent of the heat. If just a little more heat than usual has been going into the oceans, it will have had only a slight effect on ocean temperatures, because of waterās huge capacity to absorb heat, but a large effect on atmospheric temperature. And several studies suggest that recently the oceans have indeed been soaking up even more heat than normal.
Why? Well, heat constantly sloshes back and forth between the oceans and atmosphere ā this is the main cause of natural variability. What happens in the vast Pacific Ocean matters most. During a phenomenon called an El NiƱo, when westerly winds spread hot water across the top of much of the tropical Pacific, so much heat flows into the air that the entire surface of the planet warms. There was an especially strong El NiƱo in 1998, which is why it was such a warm year.
During the opposite event, called a La NiƱa, when easterly winds spread upwelling cold water across the sea surface, the tropical Pacific soaks up so much heat that it cools the planetās surface. And lately there have been lots of La NiƱas. āWe have not seen a major El NiƱo for the past 15 years,ā says Shang-Ping Xie of the Scripps Institution of Oceanography in San Diego, California (see Why didnāt models predict the slowdown?). āBut there have been several long-lasting La NiƱas.ā A model study by Xie, published in August, showed that this alone could explain the slower surface warming (, vol 501, p 403).
Whatever the cause, observations suggest that more heat than usual has gone into the ocean, and particularly the deep ocean (). āThe ocean is warming at depth,ā says study author Kevin Trenberth of the National Center for Atmospheric Research in Boulder, Colorado. So some researchers such as Trenberth and Xie think the slowdown is mostly due to the oceans.
Unknown oceans
Not everyone is convinced. The Argo network of probes for measuring ocean temperatures (see āPlumbing the depthsā) was only completed in 2007 so although we have a good idea of how much heat there is now in the oceans, it is hard to be sure how much it has changed in the past. We are also uncertain about the effects of aerosols and so on, which leaves room for debate.
The mainstream view expressed in the latest Intergovernmental Panel on Climate Change report is that about half of the surface slowdown is due to the oceans, and the other half due to the sun and extra volcanic aerosols. āItās three or four things added up,ā agrees Gavin Schmidt at the NASA Goddard Institute for Space Studies. But he is not convinced that the oceans are one of these things. They have continued to soak up heat, but we canāt be sure that they have been soaking it up faster than usual, says Schmidt.
He suspects that soaring aerosol emissions from China may have contributed to the slowdown. Possibilities like this cannot be ruled out, responds Jochem Marotzke of the Max Planck Institute for Meteorology in Hamburg, Germany, who helped write the relevant parts of the latest IPCC report, but there is no evidence for them.
While there may be uncertainty about the precise causes of the slowdown, there is no doubt about the big picture. Measurements show that sea level is rising faster than ever, at around 3 millimetres a year on average (see āDrowning by numbersā). At least 1 millimetre per year of this is due to the expansion of seawater as it warms, showing that the ocean is gaining heat independently of measurements of water temperature. The rest is due to the melting of land-based ice.
So if you define global warming in terms of the total heat content of the atmosphere, land and oceans ā as most scientists would ā then there has been no hiatus. āGlobal warming has not stopped,ā says Marotzke. āSea ice is still melting, the ocean is still taking up heat and sea level continues to rise.ā In fact, Trenberth thinks that the process may have accelerated.
Delayed action
So what happens next? Does the slowdown mean global warming isnāt going to be as bad as we thought? The fact that surface temperatures have not been rising as fast as they were is good news. To the extent that this is due to less heat coming in from the sun or more being reflected into space by aerosols, we have struck lucky. That heat is gone forever.
Increased ocean heat uptake is more of a mixed bag. Much of the heat going into the oceans will stay there. This heat will not warm the atmosphere (good) but it will contribute to sea level rise via thermal expansion (not good), and it will mean the oceans take up less heat in the future (bad).
Some of the heat now going into the oceans, though, will slosh back into atmosphere, leading to rapid surface warming (very bad). āPart of the heat is lost,ā say Trenberth. āSome of the heat comes back in the next El NiƱo.ā
The big question is when? Schmidt and others expect warming in the lower atmosphere will soon speed up again but there is no way to be sure. Most of the factors responsible for natural variability, from solar output to El NiƱos, canāt be reliably predicted. But Trenberth thinks that a longer term change in ocean conditions called the Pacific Decadal Oscillation is playing a big role. This reverses every two or three decades as a result of changing winds and if past behaviour is anything to go by, it will switch phase in the next five or 10 years, Trenberth says. If he is right, that would end the slowdown.
There is another possibility: the increased ocean heat uptake might result from ways we are altering the planet. Wind speeds have risen over the ocean, for instance. In theory, faster winds could be driving stronger vertical currents and thus pushing more heat down into the depths. If so, the slowdown could continue for years, perhaps even decades.
Assuming that isnāt the case and that surface warming will soon speed up, just how much warmer will it get? One method of estimating how much the surface will warm in response to a given rise in carbon dioxide levels ā known as climate sensitivity ā is to look at how much it has warmed in response to the rise so far. Since CO2 levels have shot up over the past 15 years but temperatures have only risen slightly, calculations using the latest figures suggest sensitivity is slightly lower than calculations done a decade ago.
Climate sceptics have seized upon this, but there are many reasons not to get excited. First, if temperatures in the lower atmosphere do rise very rapidly over the next few decades, these estimates will have to be revised upwards again. Meanwhile, other methods of estimating sensitivity, such as looking at changes in the climate further back in the past or using models, still point to higher values.
The bottom line is that talk about global warming stopping or pausing is misleading. It is good news that the worldās surface hasnāt warmed as fast as previously over the past decade. Yet we have still seen terrifying weather extremes, from unprecedented rainfall in Colorado and record heat in Australia to the power of typhoon Haiyan. All the while heat is still pouring into the oceans. All the evidence suggests that atmospheric warming will soon accelerate again, and it could do so with a vengeance.
How much has warming slowed?
The surface of the planet is now more than 0.6 Ā°C warmer than it was in 1951. : it started accelerating in the 1980s, hit 0.28 Ā°C per decade in the 1990s but fell to 0.09 Ā°C in the 2000s.
So the planet is not warming as fast as it was. But the details vary greatly according to how you calculate the trend. Pick the very hot year of 1998 as your starting point and the slowdown appears most dramatic: the rate of warming from 1998 to 2012 was just 0.04 Ā°C per decade. If this rate continued, the planetās surface would be just 1 Ā°C hotter in 2100 than in 1950 ā which is well under the ādangerous levelā of 2 Ā°C.
But these figures are based on the global surface temperature record compiled by the UKās Hadley Centre, and its record does not include the fastest-warming region on Earth, the Arctic, as there are so few observations there. According to NASAās record, which guestimates Arctic temperatures based on the nearest weather stations, the warming rate was 0.07 Ā°C per decade from 1998 to 2012. And according to , by Kevin Cowtan of the University of York, UK, which extrapolated Arctic temperatures from satellite data, the 1998 to 2012 rate was 0.12 Ā°C per decade. If he is right, warming has only slowed very slightly, from 0.18 Ā°C per decade in the 1990s to 0.16 Ā°C in the 2000s ā which would still take us over the danger limit of 2 Ā°C by 2100.
Cowtanās study is not likely to be the final word. Satellites measure temperature far above the ground, so extrapolating to the surface is difficult. But whatever the precise figures, there is every reason to think warming will not only continue but accelerate greatly over the coming century (see main story).
Why didnāt models predict the slowdown?
Prediction is not just very difficult, it is sometimes impossible. The slower rate of surface warming over the past decade seems to be due to a combination of factors, including a series of cold La NiƱas in the Pacific, an extra-low low in solar output and higher volcanic emissions (see main story). None of these kinds of natural events can be reliably predicted.
Perhaps it isnāt surprising, then, that out of 114 runs of the latest models, just three produced a trend from 1998 to 2012 as low as that observed. The reason models fail to predict short-term trends is not that the models donāt include natural variability ā they do. Individual runs of climate models zig zag up and down wildly, and often exhibit periods of a decade or more when surface temperatures barely increase or even cool despite a strong long-term warming trend, just as we are seeing now.
Rather, the problem is that the timing and magnitude of natural events in each model run differs from that in the real world. In one model run, for instance, there might be a La NiƱa in 1998, in another conditions might be neutral, whereas in the real world in 1998 there was an exceptionally strong El NiƱo.
But what if instead of allowing La NiƱas and El NiƱos to occur spontaneously, you tell the model when they really did take place? Shang-Ping Xie and Yu Kosaka at the Scripps Institution of Oceanography in San Diego, California, recently did this for the first time by feeding the recorded values for sea surface temperatures in the tropical Pacific into a model. This āhistoricalā model reproduced the observed global surface temperature from 1950 remarkably closely (Nature, vol 501, p 403).
It even reproduced many of the regional and seasonal characteristics of the slowdown. This close match was achieved even though the model didnāt include volcanic emissions after 2005 or the recent solar low. āAll this suggests to me that for the current hiatus, the Pacific [surface] cooling is the major driver,ā Xie says.
The unpredictability of natural variation means climate models may never be any good at forecasting the next five or 10 years (though some groups are trying to adapt them for this purpose). The key point is that short-term natural variability does not matter when predicting how much the world will warm over the next century or three. So the fact that models did not predict the slowdown is no reason to doubt their long-term projections. Think of it this way: we can be sure that the next winter will be much colder than the summer, even though we canāt say how temperatures will change from day to day.
This article appeared in print under the headline āThe heat is still onā
Article amended on 30 September 2019
The graphics āPutting the brakes onā and āPlumbing the depthsā have been updated since this article was first published on 4 December 2013.
