Would warming stop after greenhouse gas emissions end? Not quite

The latest IPCC report estimated our remaining “carbon budget” that would give us a chance of reaching the goal of keeping global warming below 2 degrees Celsius this century. That estimate was created using a simpler class of climate model that can crank out long or repeated simulations without tying up a supercomputer for a week. A new study using one of the most complex models, however, suggests that the simpler models get a key issue wrong: they overestimate how much carbon we can emit if we actually want to stay below 2 degrees Celsius over the centuries that follow.

A recent study used one of those simpler models (technically known as “Earth System Models of Intermediate Complexity," or EMICs) to look at how long it takes us to reap the benefits of cutting emissions. Rather than the oft-repeated estimate that there is a delay of about 40 years, the researchers found that the peak in temperature came just 10 years after emitting a simulated pulse of CO₂. However, they only modeled about a hundred years.

ETH Zürich’s Thomas Frölicher and David Paynter of NOAA’s Geophysical Fluid Dynamics Laboratory turned to a more complex model and a much longer timeframe to investigate a related question: what happens long after we stop emitting greenhouse gases?

Using their model, they simulated a scenario in which CO₂ emissions grew 1 percent each year for 100 years, raising the atmospheric concentration from 286 to 745 parts per million, and raising temperature by 2 degrees Celsius. At that point, CO₂ emissions dropped to zero. Then, they simulated nine more centuries, watching atmospheric CO₂ drop back to 476 parts per million as it gradually moved into the ocean and natural reservoirs on land.

Rather than tie up a supercomputer with an even longer simulation, they estimated the temperature after another 9,000 years using measures of the model’s characteristics and lower-resolution simulations of atmospheric CO₂ changes.

In addition, they pooled simulations of 12 similarly complex models to estimate comparable results for the same scenario. On the flip side, estimates for eight of those simpler EMIC models were also put together.

Surprisingly, the two types of models diverged significantly over the first hundred years after emissions ceased. While the simpler models hit their peak temperature at the end of the first century, cooling about 0.6 degrees Celsius by the end of the first thousand years, the more complex models kept warming. The researchers’ model warmed an additional 0.5 degrees Celsius between year 100 and year 1,000. On average, the other more complex models warmed an additional 0.2 degrees Celsius.

At the end of 10,000 years, the more complex models averaged just 0.1 degree Celsius below their temperature in year 100, while the simpler models had dropped around 0.8 degrees Celsius.

Simulations and estimates from the more complex models on the left, and the simpler models on the right, with the dark black line on both graphs showing the more complex model the researchers ran themselves.

Frölicher and Paynter, Environmental Research Letters

Because the IPCC’s “carbon budget” estimates were based on these simpler models, this has an interesting implication. If you want to stay below 2 degrees Celsius warming—not just this century, but in the centuries to come—you may need to keep to a stricter budget. You’d have to cut that budget by as much as 20 percent, in fact. Otherwise you may limit global warming to less than 2 degrees Celsius this century only to find that temperature still ticks slowly upward over the next few hundred years. That would leave future generations with the task of artificially pulling CO₂ out of the atmosphere in order to truly stabilize their climate.

Why is it that the two types of models simulate slightly different futures? The researchers have a guess. While all the models involved sit within the usual range of “equilibrium climate sensitivity”—the amount of eventual warming for an increase of CO₂ that is then held constant—the simpler EMIC models seem to get to their equilibrium faster.

The researchers think the difference comes down to the simulation of high-latitude oceans. Because there is much more mixing of shallow and deep water there, these regions of the ocean warm more slowly than water near the equator. But when these regions do warm, they have a larger impact on global temperature because of the way the feedbacks play out.

In the more complex models, it’s this gradual uptake of heat by high-latitude oceans that lifts global temperature over the centuries following the end of CO₂ emissions—this warming influence ends up stronger than the cooling influence of falling CO₂ concentrations.

In the simpler EMIC models, that isn’t true. Thomas Frölicher told Ars that he thinks that could be due to the models’ simpler cloud simulation, since clouds are an important part of the regional feedback to ocean warming.

That said, there are a number of things that remain to be nailed down here. But overall, this is another reminder that less is most definitely better when it comes to the amount of greenhouse gases we add to the atmosphere.

Open Access at Environmental Research Letters, 2015. DOI: 10.1088/1748-9326/10/7/075002  (About DOIs).