Climate models are invaluable tools for improving our understanding of the Earth’s climate system, encompassing not only its past, but perhaps more significantly, its future. Given the urgency of the climate crisis and the importance of science in addressing it, it comes as no surprise that the Nobel Foundation acknowledged the Nobel-worthy nature of climate modeling in 2021.

Climate simulations have helped determine how even small changes can have a profound impact. Take, for example, the atmospheric concentration of carbon dioxide (CO₂), which accounts for a mere 0.04% of the Earth’s total atmosphere. However small, its 50% increase from the start of the Industrial Revolution has helped explain global warming and the recent climate crisis.   

However, while invaluable tools, there are cases in which climate models fail, and the reason is quite simple: The climate is very complicated, like a machine with many moving parts, the behavior of which we don’t yet entirely understand.  

For Pietro Salvi, climate scientist at Imperial College London and the University of Reading, clouds have remained a challenging factor that have scrambled climate models for decades. In a recent study published in JGR: Atmospheres, Salvi and his colleagues sought to unravel just how clouds influence our global climate, analyzing numerous climate models and data going back to 1850. 

A climate mystery 

Much like a snowball can become an avalanche, small fluctuations in the world’s climate system can be amplified via a domino effect, creating loops that climate scientists call “feedbacks”. Clouds, in particular, warm the Earth by blocking radiation emitted by its surface, producing a positive feedback where the climate warms, more clouds are formed and amplify the response, leading to further warming and further amplification.

But clouds can also reflect sunlight away, causing a cooling effect and a negative feedback loop. It all depends on the type of clouds, where they form, and how they react to surface warming. Modeling this becomes quite tricky because cloud formation is a complicated process that depends on many variables.  

One of the regions where this cloud-based feedback has the greatest influence is the Pacific Ocean. In general, the Pacific is a key region for the global climate, but climate models are still not able to simulate the pattern effect — how ocean surface temperatures modulate global temperatures — in an accurate way.

As a result, surface temperature trends modeled in the Pacific don’t match observations, and climate scientists don’t know exactly why. Given the importance of the Pacific Ocean on the global climate, that’s a big concern.  

One at a time  

When faced with a seemingly impossible task, it is often helpful to break it down into smaller parts. This is the approach that Salvi and his colleagues took to understand how cloud-based climate feedbacks depend on the surface temperature pattern of the Pacific Ocean. 

In their study, the team used an approach called “single-forcing simulations”, in which only one climate driver or “forcing agent” is considered at a time. For example, in one simulation, only volcanic eruptions are active and their affect effect on the climate calculated, while other influences, such as greenhouse gases, solar cycles, and aerosols, are “turned off”.  

This greatly simplifies the system compared to other past modeling approaches, which consider all moving parts simultaneously, making it difficult for scientists to understand the effects of any of one component or factor. 

“The study touches on a very important concept: Forcing efficacy — a concept that quantifies how efficiently a forcing agent […] induces global warming relative to [just CO₂],” explained Yue Dong, a climate scientist at Columbia University who was not involved in the study.  “It remains an unresolved puzzle that coupled models generally fail to accurately simulate the observed [surface temperature] pattern over recent decades. There is increasing evidence suggesting that models may have biases in the forced response.” 

Comparing their model to others, the team was able to gain confidence in the accuracy of their approach. Interestingly, their initial findings indicated that climate change does not depend on solely on the concentration of greenhouse gases but appears to be related to a number of different factors, including the “strength” of the forcing agent and particular climate feedback loops, which both change over time. 

For example, their model outlined how greenhouse gases and aerosols became big influences at the beginning of the 20th century, while before this, volcanic eruptions were the dominant climate-shaping factor. Solar effects, on the other hand, remained low over the entire 170-year period they evaluated. 

The team also developed a metric they dubbed called the “climate feedback parameter”, which links a driver or forcing agent, such as CO₂, with a numerical outcome for the climate, such as warming global temperatures. This concept is called “climate sensitivity” and is integral to how accurate future predictions will be. 

“This is important for policy,” Salvi said. “Simple models […] don’t always account for different forcing agents causing different feedbacks, which could leave policymakers with wrong estimates of future warming.”  

An interesting find 

“The most relevant aspect [of the current study],” said Tim Andrews, research scientist at the Met Office Hadley Centre also not involved, “is the conclusion that volcanic eruptions potentially play such a large role in [explaining the time variations of the climate feedback loops]. This is important because […] it will impact near-term climate predictions.”  

Volcanic eruptions impact climate feedback loops by altering the surface temperature pattern of the Pacific Ocean. The fact that these climate events are so short lived and unpredictable makes it difficult to anticipate their future influence. But the team observed an interesting possibility. 

Using their model, the team were able to demonstrate that cloud-based feedback loops play the most important role in driving different global climate responses and are connected to surface temperature patterns and atmospheric stability. This is especially important in areas where rising air is stronger, such as in the oceanic region north of Australia connecting the Indian Ocean and the Pacific Ocean, which experiences the continuous formation of clouds.  

There, ocean surface water temperature is highest, so much so that it is called the “warm pool” region. The intensity of convection here depends on the surface temperature patterns via changes in the atmospheric stability. 

Using these simulated conditions, the model was able to identify a possible natural means of offsetting rising global temperatures caused by the greenhouse gas effect. “[They found that] volcanic forcings [have the] ability to efficiently cool the Indo-Pacific warming pool, which in turn efficiently drives changes in atmospheric stability and cloudiness, resulting in lower climate sensitivity,” said Andrews.  

Though the study outlines an important step in climate model development, it’s important to keep in mind that these predictions can be a double-edged sword and their results must be examined carefully.  

“It is unclear exactly how much of the […] forcing effect they have identified applies,” added Andrews. “What is clear is that more work needs to be done, and the quest for a solution to the pattern effect enigma is ongoing.”  

Reference: Pietro Salvi, Jonathan M. Gregory, Paulo Ceppi, Time-evolving radiative feedbacks in the historical period, Journal of Geophysical Research: Atmospheres (2023). DOI: 10.1029/2023JD038984.  

Feature image credit: engin akyurt on Unsplash