As scientists work to determine why some of the latest climate models suggest the future could be warmer than previously thought, a new study indicates the reason is likely to be linked to challenges simulating the formation and l evolution of clouds.
The new research, published in Scientists progress, provides an overview of 39 updated models that are part of a major international climate undertaking, the sixth phase of the Coupled Model Intercomparison Project (CMIP6). The models will also be analyzed for the upcoming Sixth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC).
Compared to older models, a subset of these updated models showed higher sensitivity to carbon dioxide – that is, more warming for a given concentration of greenhouse gases – although a few also showed lower sensitivity. The end result is a greater range of model responses than any previous generation of models, dating back to the early 1990s. If the high end models are correct and the Earth is truly more sensitive to carbon dioxide than scientists thought, the future could also be much warmer than expected. But it is also possible that updates made to the models between the last intercomparison project and this one may cause or expose errors in their results.
In the new paper, the authors sought to systematically compare CMIP6 models with previous generations and catalog the likely reasons for the broadening of the sensitivity range.
“Many research groups have already published papers analyzing possible reasons why the climate sensitivity of their models changed when they were updated,” said Gerald Meehl, senior scientist at the National Center for Atmospheric Research (NCAR ) and lead author of the new study. “Our goal was to look for any themes that emerged, especially with the high-sensitivity models. What came up again and again was that cloud feedbacks in general, and the interaction between clouds and tiny particles called aerosols in particular seem to contribute to greater sensitivity.”
The research was funded in part by the National Science Foundation, which is NCAR’s sponsor. Other supporters include the US Department of Energy, the Helmholtz Corporation and the Deutsches Klima Rechen Zentrum (the German Climate Computing Center).
Model Sensitivity Assessment
Researchers have traditionally assessed the sensitivity of climate models using two different measures. The first, used since the late 1970s, is called equilibrium climate sensitivity (ECS). It measures the increase in temperature after atmospheric carbon dioxide instantly doubles from pre-industrial levels and the model is allowed to run until the climate stabilizes.
Over the decades, the range of ECS values has remained remarkably constant – somewhere around 1.5 to 4.5 degrees Celsius (2.7 to 8.1 degrees Fahrenheit) – even as the patterns have become much more complex. For example, models included in the previous CMIP phase of the last decade, known as CMIP5, had ECS values ranging from 2.1 to 4.7 degrees C (3.6 to 8.5 degrees F) .
The CMIP6 models, however, have a range of 1.8 to 5.6 degrees C (3.2 to 10 degrees F), widening the CMIP5 spread on both the low and high ends. The NCAR-based Community Earth System Model Version 2 (CESM2) is one of the most sensitive models, with an ECS value of 5.2 degrees Celsius.
Model developers have been busy separating their models over the past year to figure out why ECS has changed. For many groups, the answers seem to boil down to clouds and aerosols. Cloud processes occur at very fine scales, which made them difficult to accurately simulate in global-scale models in the past. In CMIP6, however, many modeling groups have added more complex representations of these processes.
New cloud features in some models produced better simulations in some respects. Clouds in CESM2, for example, look more realistic compared to observations. But clouds have a complicated relationship with global warming – certain types of clouds in certain places reflect more sunlight, cooling the surface, while others can have the opposite effect, trapping heat.
Aerosols, which can be emitted naturally from volcanoes and other sources as well as from human activity, also reflect sunlight and have a cooling effect. But they also interact with clouds, changing their formation and luminosity and, therefore, their ability to heat or cool the surface.
Many modeling groups have determined that adding this new complexity in the latest version of their models has an impact on ECS. Meehl said that was not surprising.
“When you put more detail into the models, there are more degrees of freedom and more different possible outcomes,” he said. “Today’s Earth system models are quite complex, with many components interacting in sometimes unintended ways. When you run these models, you’ll get behaviors that you wouldn’t see in more simplified models.”
An immeasurable amount
ECS is supposed to tell scientists how the Earth will react to increased atmospheric carbon dioxide. The result, however, cannot be verified against the real world.
“ECS is an unmeasurable quantity,” Meehl said. “It’s a crude measure, created when models were much simpler. It’s still useful, but it’s not the only way to understand how much increasing greenhouse gases will affect the climate. “
One of the reasons scientists continue to use the ECS is that it allows them to compare current models to early climate models. But the researchers developed other metrics to examine climate sensitivity along the way, including a model’s transient climate response (TCR). To measure this, modellers increase carbon dioxide by 1% per year, compounded, until carbon dioxide is doubled. Although this measure is also idealized, it can give a more realistic view of the temperature response, at least on the shorter-term horizon of the next few decades.
In the new paper, Meehl and colleagues also compared how the TCR has changed over time since it was first used in the 1990s. The CMIP5 models had a TCR range of 1.1 to 2.5 degrees C, while that the CMIP6 model range increased only slightly, from 1.3 to 3.0 degrees C. Overall, the change in average TCR warming was almost imperceptible, from 1.8 to 2.0 degrees C (3.2 to 3.6 degrees F).
The variation in the TCR range is more modest than with the ECS, which could mean that CMIP6 models may not perform so differently from CMIP5 models when simulating temperature over the next few decades.
But even with the greater ECS range, the average value for this metric “didn’t increase dramatically,” Meehl said, rising only from 3.2 to 3.7 degrees Celsius.
“The high end is higher but the low end is lower, so the average values haven’t changed too much,” he said.
Meehl also noted that increasing the scope of the ECS could have a positive effect on science by stimulating more research into cloud processes and cloud-aerosol interactions, including field campaigns to gather better observations of how these interactions play out in the real world.
“Cloud-aerosol interactions are at the forefront of our understanding of how the climate system works, and it’s a challenge to model what we don’t understand,” Meehl said. “These modelers are pushing the boundaries of human understanding, and I hope this uncertainty will motivate new science.”