One of the most fundamental questions about climate change is also one of the thorniest: Exactly how much will the earth warm in response to future greenhouse gas emissions?
The answer, say scientists, is in the sky above our heads. Clouds are the fluffy, unlikely gatekeepers of climate change – they play a crucial role in how quickly the world is warming.
A number of recent studies have shed new light on this role. As the world warms, cloud cover will change around the world. And these changing clouds are likely to accelerate global warming.
That means the earth may be a little more sensitive to greenhouse gases than some older estimates suggested.
“Clouds are a great uncertainty,” says Paulo Ceppi, climate scientist at Imperial College London and co-author of one of the new studies. “And that was the main motivation. We want to understand how clouds change and how this cloud feedback affects global warming. “
Cloud research is a tricky business. Clouds sometimes have a warming effect on the local climate and sometimes a cooling effect – it all depends on the type of clouds, the local climate and a variety of other conditions.
Climate change only complicates matters. Global warming is expected to increase certain types of clouds in certain locations and decrease them in others. All in all, it’s a large, complex patchwork of effects around the world.
For years, scientists have struggled to determine exactly how clouds would change with future warming – and whether they could make climate change worse or dampen some of its effects. It was a difficult question to answer. Scientists usually use computer models to make predictions about future climate change. But clouds are notoriously difficult to simulate, especially on a global scale.
In the past few months, however, several studies have started to get to the bottom of this. They all come to the same conclusions: some of the worst global warming scenarios may be less likely than scientists previously thought. But some of the best-case scenarios will certainly not materialize either.
These studies all revolve around the same question: how much, exactly, would the world warm up if carbon dioxide levels in the atmosphere reached twice their pre-industrial levels?
It’s a hypothetical question for now. But that could change soon.
Before the industrial revolution, about 150 years ago, global carbon dioxide levels were around 280 parts per million. Double that would be 560 ppm. Today the concentrations are already above 410 ppm and are increasing every year.
This CO2-The doubling question – a metric known to scientists as “equilibrium climate sensitivity” – has been a central question among climate researchers for decades.
It was also difficult to make progress.
In 1979, a seminal report from the National Academy of Sciences suggested that the planet would likely warm by 1.5 to 4.5 degrees Celsius in response. For years one study after another came to more or less the same result.
Only recently have researchers started narrowing it down – and improvements in cloud research have a lot to do with it.
Last year, a groundbreaking new study found that CO. Emissions are doubling2 would likely lead to a warming of 2.6 to 3.9 degrees Celsius.
It’s a much narrower projection that eliminates some of the higher quality projections and eliminates much of the lower area. The study summarized the latest research on climate sensitivity and considered several different lines of evidence – including recent advances in cloud research.
And in the past few months, several recent studies – mostly focused on clouds – have also supported a narrower range of climate sensitivity.
A February study in Nature climate change suggested a likely sensitivity of about 3.5 ° C. A May study, also in Nature climate change, she put it at around 3 ° C. Both studies suggested that clouds on a global scale would likely have a moderate amplifying effect on the rate of global warming.
These studies used real world observations to draw their conclusions. They collected large amounts of data on cloud behavior – how clouds react to changes in temperature, humidity, and other weather variables – and then performed statistical analyzes of these observations to find out how clouds are likely to respond to future climate change.
According to Mark Zelinka, climate scientist and cloud expert at Lawrence Livermore National Laboratory and co-author of both the May study and last year’s study, this is a fairly traditional way of addressing the problem.
A more recent study, on the other hand, takes a less conventional approach. Published last week in Proceedings of the National Academy of Sciences, the study used machine learning to find out how clouds react to changes in their environment.
Machine learning is a branch of artificial intelligence in which computers sift through large amounts of data, recognize patterns and then use these patterns to construct algorithms that predict how future data should behave under different conditions. In this case, the researchers used real-world observations of how clouds react to environmental changes.
The machine learning approach came to a similar conclusion: a tighter climate sensitivity that rules out most milder climate scenarios. The study found that there is almost no chance of climate sensitivity below 2 ° C.
“For some time I have thought that the cloud problem is particularly suitable for machine learning approaches,” says Ceppi, who carried out the study together with his climate scientist and machine learning expert Peer Nowack. “If you want to understand the relationship between clouds and temperature or humidity or wind, it is quite difficult to tease out the individual effects of each of these environmental variables.”
Machine learning can be an easier way to tackle such a complicated dataset, he said.
Machine learning shows promise in other types of cloud research as well. Some research groups are experimenting with integrating machine learning components into global climate models in order to circumvent the difficulties involved in simulating clouds.
Clouds pose a challenge to models because they require extremely fine physics – after all, clouds form tiny water droplets in the sky. The simulation of these microscopic processes on a global scale would require unimaginable computing power; it’s just not possible.
To get around this, modelers typically don’t force their models to physically simulate cloud formation. Instead, they manually insert information about how clouds should form and respond to changes in their environment, a tactic known as parameterization.
Machine learning can be an alternative to parameterization. Instead of introducing a rule about how clouds should behave within the model, a machine learning component can construct algorithms that predict how the clouds should react.
It’s not exactly a common strategy yet. But several research groups have started looking into how useful it could be in recent years.
These are promising advances in the intricate field of cloud research. Still, “machine learning is a very helpful tool, but not a panacea,” warned Piers Forster, director of the Priestley International Center for Climate at the University of Leeds, in an email to E&E News.
Machine learning is an efficient way to analyze complicated data sets – but it can leave some questions about the underlying physical processes behind that data unanswered. There is still plenty of room for more traditional research into the how and why of cloud behavior.
“In my opinion, coordinated developments on both fronts are the answer,” added Forster.
In the meantime, Zelinka added, it is reassuring that different strategies have produced similar results.
“If it were just a study, the robustness of this result could be questioned,” said Zelinka. “But when you have more and more evidence from independent authors using independent techniques and they all come to a similar conclusion, that’s pretty powerful.”
Reprinted from E&E News with permission from POLITICO, LLC. Copyright 2021. E&E News provides important news for energy and environmental professionals.