Some climate models predict future warming that most experts don’t believe will be credible. Image source: UGURHAN/ISTOCK
One study suggests that Arctic rainfall will dominate in the 2060s, decades earlier than expected. Another report said air pollution from forest fires in the western United States could triple by 2100. There is also a theory that a marine mass extinction could come within centuries.
All three of the above studies, published last year, rely on projections of the future by some of the world’s next-generation climate models. But even modelmakers acknowledge that many models have an obvious problem: predicting a future that is too hot to overheat. Some researchers worry that this has led to a series of “faster than expected” results that could undermine the credibility of climate science.
Recently, climate scientists wrote in a commentary in Nature that researchers need to be more “picky” about how to use model results. Researchers should no longer simply use the averages predicted by all climate models, which could lead to global temperatures being 0.7 degrees Celsius higher by 2100 than the Intergovernmental Panel on Climate Change (IPCC) estimates.
Lead author Zeke Hausfather, head of climate research at stripe, a payments services company, said, “We have to get rid of the naïve idea of a democratic model. Instead, he and his colleagues called for an elite model, sometimes prioritizing the outcomes of models known to have more realistic rates of warming.
Co-author Kate Marvel, a climate scientist at NASA’s Goddard Space Flight Center, said, “The ‘overheating’ model doesn’t invalidate the principles of climate science, and the greenhouse effect is still warming the planet.” In her opinion, these models are not perfect, “they are not crystal balls”, but in general they are still very successful research tools.
In 2019, the International Comparative Scheme for Coupled Models (CMIP) raised the issue of model overheating. The program synthesizes the results of global models before the IPCC publishes a major report every 7 or 8 years.
In previous rounds of CMIP, most models predicted “climate sensitivity” — projecting warming of 2 to 4.5 degrees Celsius when atmospheric carbon dioxide levels doubled from pre-industrial levels. But in 2019’s CMIP6, 10 of the 55 models had sensitivities above 5 degrees Celsius — a significant deviation.
This result is also inconsistent with a landmark study that eschewed global modelling and instead relied on paleoclimate and observational records to determine Earth’s climate sensitivity. The study found that the value was between 2.6 and 3.9 degrees Celsius.
Christopher Field, a climate scientist at Stanford University in the United States, said the estimated difference in sensitivity is “a sobering example of the complexity of the climate system.”
Since then, researchers have been tracking the causes of the overheating model. These models include models produced by the National Centre for Atmospheric Research, the U.S. Department of Energy, the British Weather Service and the Canadian Centre for Environment and Climate Change.
In fact, many of these models are better than their “predecessors,” Marvel said, and model makers have been open to the issue and “deserve credit.” But it will take years for them to come up with a widely used new predictive model.
In 2021, the IPCC released its first working group report, which covers the physical basis of climate change, in an attempt to remedy the problem. The IPCC rates the model based on its ability to capture historical temperatures, and then uses the model to make an official “assessed warming” forecast for different fossil fuel emission scenarios.
In studying future changes on earth, the IPCC reported results based on models based on all degrees of warming: 1.5 degrees Celsius, 2 degrees Celsius, 3 degrees Celsius. This allows useful information for overheating models to be exploited, even if they reach thresholds too quickly.
Hausfather himself is a co-author of the IPCC. He said that while the IPCC accepted the challenge, it didn’t do a good job of telling everyone what the real problem was.
Since then, dozens of published studies have used predictions based on the original averages of all CMIP6 models, and the results tend to be “worse” than the IPCC’s predictions — which has caught the attention of those who don’t understand the underlying problems with the model. Marvel said, “It’s not because someone has malicious intentions. It’s just because there’s no guidance. ”
The authors say climate impact researchers need to follow the steps taken by the IPCC. First, avoid dubious time-based scenarios that emphasize the effects of specific levels of global warming, regardless of when those levels were reached. Second, use the IPCC’s own “Assess Warming” forecast to predict when the degree of warming is likely to occur. For studies where the details of the warming trajectory are important, they can use models that capture warming relatively accurately, such as those made by agencies such as NASA and the National Oceanic and Atmospheric Administration.
Clautia Tebaldi, a climate scientist at Lawrence Berkeley National Laboratory in the United States and one of the leaders of CMIP’s climate prediction program, said, “I agree with almost everything the author says and suggests.” However, she also notes that the recommendations may underestimate decision-makers’ thirst for time-based information, not just the absolute amount of warming.
Reto Knutti, a climate scientist at ETH Zurich in Switzerland, said, “Researchers should consider further scenarios and examine some models for large regional biases.” (Source: China Science Daily Wang Fang)
Related paper information:https://doi.org/10.1038/d41586-022-01192-2