The computer models that predict climate change may be overestimating the cooling power of clouds, new research suggest.
If the findings are borne out by further research, it suggests that making progress against global warming will be even harder.
The new paper, in the journal Science, focuses on the balance of water and ice in clouds, which affects the impact that carbon dioxide levels have on atmospheric temperatures, a factor known as equilibrium climate sensitivity.
The research suggests the effects of a flaw in the model could be serious: based on its analysis of one model of climate change, the cloud error could mean an additional 1.3°C of warming than expected.
The findings, if proved correct, would vastly reduce the amount of carbon that could be emitted to achieve the Paris target of 1.5°C warming.