Caltech researchers use AI to forecast climate tipping points in new study
A Caltech-led team published a paper in the Proceedings of the National Academy of Sciences demonstrating that an AI program can predict when simple climate models reach irreversible tipping points.
New AI tool predicts system shifts
The study utilizes a neural network called a recurrent neural operator to identify moments when systems become unstable. This technology applies to simulations of climate change and turbulent airflow over aircraft wings. Scientists previously used similar models only for smaller weather variations like daily temperature changes.
Definition of irreversible change
Principal author Anima Anandkumar describes a tipping point as a state where a system cannot return to its original condition. Once parts of the system change in this way, the process becomes nonstationary and potentially irreversible. Examples cited include global warming and the melting of polar ice caps.
Publication details and scope
The research paper appears in the Proceedings of the National Academy of Sciences under the title Forecasting tipping points with AI climate simulation. The team successfully tested their method on models representing both environmental systems and aerodynamic airflow. This approach addresses the higher uncertainty involved in forecasting large-scale shifts compared to standard weather predictions.
Reported by one outlet
Only one outlet has published this. Nothing here has been checked against a second report, so read it as that outlet's account and follow the link below for the original.
Reported by
1 independent outlet. Headline as published. Links open the original report.