Science Climate Science

What really happened during the supposed climate change 'hiatus'?

Researchers have independently replicated NOAA's recalibration of sea surface temperature data to uncover what really happened from 1998 to 2012.

A child plays in a fountain in Erfurt, central Germany, on a hot day in September 2016.
Jens Meyer/AP/File | Caption

Nope, climate change didn't pause for about 15 years, scientists say – again. But just how did that misunderstanding happen?

From around 1998 to 2012, the rise in global temperatures seemed to plateau, according to NOAA's Extended Reconstruction Sea Surface Temperature (ERSST) dataset.

To most climate scientists, this so-called hiatus was another puzzle of our complex climate system for them to work out. But to people who were already skeptical of global warming, this data was evidence for the idea that human-induced climate change is a hoax.

But the data itself was unsound, scientists now say. There is no evidence of a hiatus. 

Climate scientists use sea surface temperature data in their calculations of global warming trends. But "a fair bit of the apparent hiatus seems to be due to problems in our ocean measurements, and not a real thing," as study lead author Zeke Hausfather, an energy systems analyst and data scientist at the University of California Berkeley and Berkeley Earth, tells The Christian Science Monitor in a phone interview.

Dr. Hausfather and his colleagues aren't the first team to say this. But in 2015, when National Oceanic and Atmospheric Administration (NOAA) scientists first identified the errors in the data, political turmoil ensued. So, as replication is a tenet of science, Hausfather and his colleagues set about seeing if they could come to the same conclusion as the NOAA team. Their results are detailed in a paper published Wednesday in the journal Science Advances.

"We showed that NOAA was right," Hausfather says. "Despite all the flack that they got for their study, they were doing good science and their results agree with all of these other independent and instrumentally homogenous temperature records that we looked at." 

In 2015 a team of NOAA scientists published a revision of the ERSST data in the journal Science, showing that there had never been a pause in global warming. The previous version, ERSST version 3b, had a "cool bias" over the period of 1998 to 2012, showing temperatures as lower than they should have been, and therefore, they wrote, "the slowdown was just an illusion." 

So just what caused the data to skew cold? Changing technology, the NOAA scientists and Hausfather agree.

For decades, sea surface temperatures were measured on ships, using water sucked into a ship's engine room. But in the mid-1990s, scientists began deploying a new strategy across the world's oceans: thermometers on buoys. By the late 1990s, this method had taken off, Hausfather says.

But here's the catch: the buoys bobbing on the surface of the sea take colder measurements than those taken within the warm engine room of a ship. And, in the ERSST version 3b, scientists had just tacked on the buoy data without adjusting for that difference. 

When NOAA scientists realized the problem, they calculated that difference and weighted the data differently, resulting in ERSST version 4. And, according to their calculations, version 4 showed more than twice as much warming, on the global scale, as version 3b.

That 2015 NOAA paper triggered political controversy.

Some politicians, led by Rep. Lamar Smith (R) of Texas, who chairs the US House Committee on Science, Space, and Technology – and whom critics call a climate denier – pushed for an investigation into the NOAA study, suggesting that the scientists had manipulated the version 4 data to get politically motivated results.

To investigate that period of time independently, Hausfather and his colleagues decided to look at the data a little differently than the NOAA scientists had. Most scientists look at composite sea surface temperature records, he explains, whether it's the ERSST from NOAA, HadSST3 from Britain's Hadley Centre, or COBE-SST from the Japanese Meteorological Agency. These composite datasets are put together with data collected from multiple sources, as in both ships and buoys. 

Instead of looking at all the data mashed together, Hausfather and his colleagues studied the trends in the data from different sources separately, including data from ships, buoys, satellites, and drifting robots called Argo floats. 

"The authors did a great job of inter-comparing independent and semi-independent SST data sets," Thomas Karl, the lead author on the 2015 NOAA paper and former director of NOAA's National Centers for Environmental Information, writes in an email to the Monitor. "Their results show the importance of independent measurements to clarify observational uncertainties."

"This paper further allays any qualms that there may have been scientific errors or any non-scientific agendas," Kevin Trenberth, senior scientist in the Climate Analysis Section at the National Center for Atmospheric Research, who was not part of either paper, writes to the Monitor. "Lamar Smith owes Tom Karl an apology." 

Moreover, the perceived significance of a "hiatus" was itself a misunderstanding, says climatologist Gavin Schmidt, the director of the NASA Goddard Institute for Space Studies, who was not involved in this study. "If you think that very small changes on a very short number of years makes a difference to the big picture, you're just fooling yourself," he tells the Monitor in a phone interview. 

The climate system is complex and can vary naturally over a decade or two, he explains. But that is separate from longer-term trends. 

"It must be emphasized that there are (at least) two main things going on in the climate system. One is the long-term trend due to increasing [greenhouse gases], and the other is decadal-timescale naturally-occurring internally-generated climate variability," writes Gerald Meehl, senior scientist at the National Center for Atmospheric Research, who was not involved in this study, in an email to the Monitor. "And this is not an either/or situation – BOTH are important to understand what we experience in the climate system."

"So to say the slowdown never occurred is to ignore the important aspects of internal variability," he explains, "and to say that global warming stopped in the early 2000s ignores the important long-term warming trend due to increasing [greenhouse gases]."

And as the past few years set temperature records, more evidence is mounting that global warming has not stopped, both Schmidt and Hausfather say.

That's not to say that nothing strange happened in the temperatures on Earth from 1998 to 2012. Even in the researchers' adjusted calculations, temperatures during that 15 year period did seem to be rising a bit slower than models had predicted. And that is why scientists have been fascinated by this period of warming: it could help refine climate models.

What might have driven the slowdown?

"There were a lot of little things that happened around the same time," Dr. Schmidt says. 

For example, there was a large El Niño event right before that 15-year period that would have made the following years seem comparatively cold. One study suggested that the other side of the El Niño cycle, La Niña, could have also contributed to those lower temperature readings.

In addition, some small volcanoes that hadn't been accounted for in climate models could have contributed some to the cooling effect, Hausfather says.

"There still is a lot to focus on, from a scientific standpoint, in terms of better understanding short-term variability in that period," he says, but "I don't think it really changes our understanding or expectation around long-term warming. There's no evidence to date, for example, that that particular period of slowdown means the climate's not as sensitive to CO2 as we thought, or that the models that show less warming are right."

Still, Hausfather says, "We have to constantly be vigilant to potential biases in our understanding.… There is no perfect measurement system, particularly over long periods of time."