'The Pause' vs. 'The Surge'

What to think about this year’s very rapid warming

This summer has been incredibly hot, as Zeke wrote a few weeks ago. But then September came along and its extreme warmth was so astonishing that plots had to dramatically expand their y-axis to get the month in. Climate scientists, rarely a group with nothing to say, have been at a loss for words to describe it. Zeke called it “gobsmackingly bananas,” a term that seems to have stuck.

Globally averaged surface air temperature anomalies relative to 1991–2020 for each September from 1990 to 2023. Data source: ERA5. Copernicus Climate Change Service/ECMWF.

I’ve been getting a lot of calls from reporters asking the same thing: what’s going on? Have we reached a tipping point? Have we broken the climate?

The pause

In one way, this feels like deja vu from the early 2010s, when climate deniers were saying, “Global warming has stopped!”

article

As with most denier arguments, it was based on a nugget of truth: the trend in global average surface temperatures between the late 1990s and mid 2010s was indeed close to zero, a phenomenon that became known as the pause or the hiatus.

Deniers implied that this somehow challenged climate science, but the scientific community always knew that warming would eventually resume. The oceans were accumulating energy over this entire time, so it eventually had to appear at the surface as increasing temperatures. The only question was when.

It was also obvious how you can claim no warming over short periods (e.g., a decade) while the climate was actually warming. To understand this, you can think of the global temperature record as the sum of a long-term trend, driven by the increase in carbon dioxide and other greenhouse gases, and short-term variability from things like El Nino-La Nina cycles:

the global temperature record can be thought of as the sum of a long-term trend and short-term variability

If you add these components up, you get something that’s (at least schematically) similar to the observed temperature record.

simple trend + sin wave in left-hand panel and observed global average temperature from satellites in right-hand panel.

Once you see this, it’s easy to get a zero or even negative trend in a world that’s warming — you just carefully pick your start year to be hot and your end year to be cold:

you can get a “cooling trend” in a warming climate by careful selection of your start and end years.

So when Ted Cruz says in 2015: “… satellite data demonstrate for the last 17 years, there’s been zero warming,” you can see what’s he’s doing. He’s starting in a very hot year (1998) and ending in a cool year (2015):

Data analysis Ted Cruz style! UAH satellite monthly temperature anomalies are plotted.

This is supremely dishonest data analysis. But the more important fact is this: by judiciously picking your endpoints, you can almost always argue that global warming has stopped:

from skeptical science

The pause ultimately ended in 2016 when temperatures associated with an El Nino smashed previous records — and the Earth resumed warming¹. When looking at the long-term temperature record today, it agrees well with what the climate models have predicted for the last few decades.

At this point, I’m reasonably sure that the pause was a big nothing. It was probably just random variability of the climate system, perhaps combined with a few other factors, like slightly reduced radiative forcing due to several small factors. There’s nothing there to challenge our understanding of the climate system or projections of the future.

The surge

As a counterpoint to the pause, let’s call the extreme temperatures of this year ‘the surge’. Like the pause, people are looking for long-term lessons from a short snippet of the climate record. That’s dangerous: as we saw with the pause, short-term climate variability can mislead you about the future of the climate.

This is not to say that things are not changing. If you look at the ensemble average of CMIP6 models, we see that the average trend over 1970-2014, 0.22C/decade, increases to 0.30C/decade over 2015-2030².

Given how good the climate models have been at predicting global average surface temperature, we ought to give significant deference to this prediction. I remain open to arguments that the models are underestimating (or overestimating) future warming, but the burden of proof is on those trying to make the case.

Ultimately, nature will settle this just as it settled the hiatus. If we continue to see extraordinary warmth for the next 3-5 years, even after the current El Niño ends, we’ll know that something fundamental has changed in our climate system. However, I suspect that things will settle down and the climate will continue evolving along the pathway charted by the climate models.

This provides me no comfort, though. My biggest fear is not that the physical climate system spins out of control, but that we reach tipping points in the economic and social systems due to climate change, such as an implosion of the insurance market. The impacts of climate change are strongly non-linear, so our society could be much closer to a catastrophic tipping point than we realize. That’s what keeps me up at night.

1. Amusingly, climate deniers now say, “There’s been no warming since 2016”. Without doubt, in the future they’ll be saying, “There’s been no warming since 2023”.

2. Thanks to Zeke for calculating these numbers.