For years, the global warming ‘hiatus’ from 1998 to 2012 puzzled scientists and fueled skeptics looking to cast doubt on the very idea that Earth’s temperature has been on the rise, largely because of human-produced greenhouse gas emissions such as carbon dioxide — and that significant policy changes would need to be made to keep that rise in check.
Recent papers have begun to chip away at the idea of the slowdown. Now, a new analysis in the journal Nature brings together many of those arguments to show that the hiatus may not have been quite what it seemed.
“A combination of changes in forcing, uptake of heat by the oceans, natural variability and incomplete observational coverage reconciles models and data,” the researchers from the Institute for Atmospheric and Climate Science in Switzerland wrote. “Combined with stronger recent warming trends in newer datasets, we are now more confident than ever that human influence is dominant in long-term warming.”
Below, we sort through a few of the issues with the hiatus — and the lessons scientists learned from it.
What was the hiatus, anyway?
That’s a good question. Part of the problem is that there doesn’t seem to be an agreed-upon definition. The hiatus has been defined in three main ways in the scientific literature:
- A period in which there is no significant positive trend in global mean surface temperatures (when it’s essentially flat)
- A shorter-term slowdown in rising temperatures compared to the preceding long-term warming trend
- The rate of increase in temperatures is lower than scientific models predicted
So whether you call it the ‘hiatus’ or the ‘pause,’ both are arguably misnomers, depending on which meaning is used. After all, in the latter two definitions, temperatures are generally still rising — just not as much as expected. And depending on which definition you use, different models will be more accurate than others.
“All three of those are very different arguments …. so it often makes it fairly confusing to talk about this hiatus discussion,” said Zeke Hausfather, a climate scientist at Berkeley Earth and PhD student at UC Berkeley who was not involved in the paper.
“The only sort of ‘true’ definition of a hiatus would be if global warming actually stopped,” he added, “and there’s no evidence in any operational datasets today that that happened.”
Regardless of the definition, to whatever the extent the global warming slowdown existed, it’s definitely over now. Independent analyses by NASA and NOAA show that 2016 was the hottest year on record, marking the third year of record-breaking heat in a row. The causes remain linked to human-produced greenhouse gas emissions.
So the hiatus is over? Then why do scientists care?
Scientists had to care because everyone else seemed to. The hiatus, which is marked from 1998 to 2012, was generally treated by climate scientists as a reflection of short-term variations that didn’t affect long-term trends.
“In leaked drafts of the Intergovernmental Panel on Climate Change (IPCC) Working Group I (WG1) Fifth Assessment Report (AR5), the global warming hiatus was considered to be consistent with natural variability, and hence not in need of a detailed explanation,” the study authors wrote. “At the time of the first draft, there was almost no literature on the hiatus to be assessed anyway. Scientists knew from observations and models that global temperatures fluctuate on timescales of years to decades.”
But the deviation from expectations was seized upon by the media as well as global warming skeptics, the study authors noted, turning it into a political football around the time that major policy decisions about how to deal with climate change were under discussion.
“The interest of the media and public grew, and groups with particular interests used the case to question the trust in both climate science and the use of climate models,” the authors wrote.
(The Times has reported on fossil fuel companies’ support of climate skepticism.)
Those short-term trends may be considered relatively minor fluctuations — but they matter on human timescales, and they’re still poorly understood.
“Now, in 2017 … after a wave of scientific publications and public debate, and with GMSTs [global mean surface air temperatures] setting new records again, it is time to take stock of what can be learned from the hiatus,” the authors wrote.
It is time to take stock of what can be learned from the hiatus.
— Authors of the Nature study
So what did this new analysis find?
This analysis in Nature pulls together many findings in the wake of several research efforts looking to address the hiatus conundrum. Some argued that global warming predictions didn’t match the actual data during that time period because they didn’t include complex short-term climate factors that are poorly understood. Others have argued that the hiatus didn’t really exist; that it was a problem with the instrumentation or the data analysis, for example.
The new Nature paper essentially works through several examples under both of these explanations and finds that it’s a combination of the two. For example:
- Regions in the Arctic, where global warming is having marked effects, are actually sparsely monitored, and different analyses try to extrapolate from that limited data set in different ways.
- The definitions of a pause are faulty, especially when you consider that “the data continue to show significant warming trends when the trend length exceeds 16 years,” James Risbey and Stephan Lewandowsky, who were not involved in the paper, wrote in a commentary on it. (Risbey is with the Oceans and Atmosphere Commonwealth Scientific and Industrial Research Organisation in Australia; Lewandowsky is from the University of Bristol.)
- The observed global mean temperature data takes into account both ocean and surface air temperatures, but model predictions have often only used air temperatures. This leads to apples-and-oranges comparisons between predictions and reality, Risbey and Lewandowsky said.
- Short-term ocean dynamics patterns, such as those of El Niño, the Pacific Decadal Oscillation and the Atlantic Multidecadal Oscillation, move and store heat in ways that lead to surface temperature fluctuations that can last years or even decades. This does not mean that the heat’s gone, just that it’s out of range of sensors. And because we don’t fully understand those patterns, and when exactly one will start and when it will end, short-term predictions can be a little off — even though they’re right in the long-term.“The models have their own El Niño events but they’re not necessarily happening at the same time as the real-world El Niño events,” Hausfather said.
- Climate change models did not perfectly characterize all the right complicating factors — how much solar radiation there was in a given year, for example, or the extent of the cooling effect from volcanic activity and human-produced aerosols. When those estimates are updated to reflect what actually happened, the models get much closer to reality.
- Recent studies showed the need to better account for temperature detection changes that happened as ocean temperature monitors switched from ships to buoys. The buoys may have made temperature readings look a little cooler than expected, even when they were not. (In the older method, water would be heated by the ship’s engine on its way to be measured, while the buoys would measure water temperatures in the open ocean.)
“If you do all of these things you end up getting observations that match models pretty exactly,” Hausfather said of all the mitigating factors included by the Nature study authors. “Each of these things is a little part of the divergence between models and observations.”
Another problem? Marking the beginning of the hiatus from 1998, which was a year of record-breaking heat.
“Picking a large El Niño event as the starting year, you’re necessarily going to have a lower rate of warming after that, even in a warming world,” Hausfather said.
The problem, ultimately, is that studies looking at long-term change simply aren’t especially good at predicting short-term variation. That doesn’t mean that the long-term trend isn’t there; human-caused carbon dioxide emissions continue to fuel global warming.
“In short, some data, tools and methods that were good enough when looking at longer-term climate change proved to be problematic when they were focused on the problem of explaining short-term trends,” Risbey and Lewandowsky wrote. “Small differences in GMST data that are inconsequential for climate change are amplified when short-term trends are calculated. Climate-model projections are blunt tools for the analysis of short-term trends.”
What new light does this paper shed on the hiatus?
This analysis isn’t a “bombshell,” pointed out Michael Mann, a climate scientist at Penn State University.
“The work of many groups (including our own) has shown that models and observations are consistent in terms of long-term warming, and that this warming — and recent extreme warmth — can only be explained by anthropogenic (human-caused) activity, namely the burning of fossil fuels,” Mann said in an email. “The fact that there is substantial internal variability that can mask this warming on decadal and multidecadal timescales is also something established by us in previous work.”
But it does offer a holistic explanation of all the research that has been chipping away at this climate conundrum, Hausfather pointed out.
“I think it does a really good job at tying a bow on the last five years or so of hiatus arguments,” he said.
Is there a lesson to be learned from the controversy around the hiatus?
The study authors seem to think so. For example, scientists were forced to better understand these short-term variations, which helped them pin down factors such as how sensitive the climate was to additional carbon dioxide, the dominant greenhouse gas in Earth’s atmosphere.
“As a consequence, after a surge of scientific studies on the topic, we have learned more about the ways in which the climate system works in several areas,” the authors wrote.
And then there are other evergreen lessons — particularly with respect to science and its interface with public discourse.
“Social sciences might find this an interesting period for studying how science interacts with the public, media and policy,” they wrote. “In a time coinciding with high-level political negotiations on preventing climate change, sceptical media and politicians were using the apparent lack of warming to downplay the importance of climate change…. This will not be the last time that weather and climate will surprise us, so maybe there are lessons to be learned from the hiatus about communication on all sides.”
Already a subscriber? Thank you for your support. If you are not, please consider subscribing today. Get full access to our signature journalism for just 99 cents for the first four weeks.
MORE IN SCIENCE