(Bloomberg) — Something funny happened when, three years ago, climatologists started reporting results from their latest generation of Earth models. A disturbing number of them showed accelerated heating.
These “hot models” quickly sounded the alarm. It’s a puzzle that is now well on its way to being solved, according to a commentary published today in the journal Nature by five leading scientists. Advances that include a better understanding of cloud physics have had the unintended effect of increasing global temperatures faster in some models.
With the effects of climate change visible virtually every week, governments, businesses, investors and communities are all hungry for high quality information on what’s to come. How to deal with hot patterns is a key part of this, and the authors outline how to tap into what’s useful about them.
Climate models are the workhorses of complex research that for decades have, for the most part, correctly projected how temperatures would rise due to continued greenhouse gas pollution – “the best window on the world future that climatologists have,” said Richard Betts, chief climate officer. impact research at the UK Met Office, which was not involved in the new article. They’re driven by the physics of how carbon dioxide and other gases absorb heat, how much light ice caps reflect back into space, and a myriad of other phenomena that generations of scientists have observed. in the real world.
Where physics can’t fully capture natural processes, modelers are getting closer to the behavior of the world – and that’s probably where the headache of the latest model updates has come. These models, which formed the backbone of the latest United Nations climate assessment that began publishing in August, included more sophisticated looks at the science of ice, water and clouds that seem pushing some model answers to the central question of how much CO₂ causes how much warming.
The United Nations Intergovernmental Panel on Climate Change took the problem of hot models seriously in its latest series of reports and ended what scientists call “model democracy”, or the treatment of each search pattern as equal to all others. Hot models received less attention, and the IPCC added another layer of analysis afterwards.
Despite recent work by the IPCC, the Nature authors fear that in the coming years, it’s nobody’s job to encourage best practices that keep hot patterns from showing up in research where they could indicate more warming than what is already taking place. They point out that what is expected of climate change is bad enough, and as governments and businesses increasingly examine what climate projections mean specifically for them, the most accurate science is more critical than ever.
Overall, the output of the most recent models “represents a lot of the data that will then be used by businesses, local governments, local decision-makers for adaptation and planning purposes,” said Zeke Hausfather, author and head of climate research at Stripe. “The last thing we want to do is tell everyone that we are going to have a lot more warming than expected.”
The first of the Nature article’s largely technical recommendations regarding hot models is useful beyond the pages of scientific journals. It’s about starting to think about climate change more in terms of levels of warming than in what year we might reach them. When the rate of warming, which is the central problem of warm models, is not a key factor in new research and in particular on risks, said Betts, the models are still useful to “provide the most complete information on potential impacts and adaptation needs”.
Climate change is inherently full of surprises, which makes timelines – what will happen by 2030, by 2050, by 2100 – less reliable than understanding what each increasing temperature rise means undesirable – 1.5°C, 2°C, 3°C.
“The bottom line is, well, what’s the best optimal time to hit 2° warming? It’s never, is it? Let’s not do that,” said co-author Kate Marvel. of the article and research scientist at Columbia University and NASA’s Goddard Institute for Space Studies.
©2022 Bloomberg LP