Theoretical work confirms the method used to predict climate tipping is indeed mathematically sound
Recently, scientists from the TiPES project have found early warning signals for climate tipping of the Amazon Rainforest, the West-Central Greenland ice sheet, and the Atlantic Meridional Overturning Circulation (AMOC, a large-scale heat-distributing ocean current system ). It has not, however, been absolutely clear to which degree the heuristic approach behind early warning signals might lead to false alarms. Now, improvements in the theoretical understanding of tipping points confirm that early warning signals are indeed robust indicators of a climate system closing in on a critical transition, leading to a changeover in its behaviour. The study by Manuel Santos Gutièrrez, the Weizmann Institute, Israel and Valerio Lucarini, the University of Reading, United Kingdom is published in Journal of Physics A
Early warning signals for abrupt transitions (tipping points) are being intensively studied because such signals might give us time to prevent or mitigate the consequences of a climate tipping event. The kinds of signals, that are considered signs of a climate system approaching a tipping point are increased variability, increased sensitivity and critical slowing down.
Wavy and slow
To illustrate these concepts, imagine a chair being slowly tilted towards the point where it is balancing on two legs and could fall to either side. When it is close to this point, it will 1) react more intensely to a small push, (increased sensitivity), 2) exhibit larger fluctuations (increased variability) and 3) take longer to fall back to the point it was before it was pushed (critical slowing down). In other words, small perturbations to the system seem to be amplified and the system becomes more wavy and slow.
All physical systems that have two or more stable states (for the chair: lying or standing) behave like this before they tip from one state to another. The climate system, however, is overwhelmingly complex. So it has not been exactly clear if increased variability, increased sensitivity, and critical slowing down might always be considered signs of a tipping point approaching.
Indeed the same physics
The theoretical work presented in the article On Some Aspects of the Response to Stochastic and Deterministic Forcings now delivers a clear mathematical understanding that increased variability, increased sensitivity, critical slowing down, and being close to a tipping point, are indeed part of the exact same physics.
”This provides stronger support for understanding how early warning systems work and gives an indication of how to improve them. It also clarifies why machine learning methods for predicting tipping points work,” explains Valerio Lucarini.
”In particular we understand now better why combining different time series describing the same system improves our early warning signal method, and it shows that we can use the same mathematical tools to perform climate change prediction, to assess climatic feedback and indeed to construct early warning signals,” says Lucarini.
All in all, the paper provides a strong foundation for data analysis and interpretation in the whole area of Tipping points.
Must be taken seriously
Early warning signals for tipping have in recent years been found in three major climate subsystems. In 2022, early warning signals were found in data from the Amazon rainforest.
And in 2021, data from the ice sheet of the Western Central part of Greenland and the AMOC were both shown to bear telltale marks of approaching tipping points. If any of these three large subsystems of the Earth system tip, consequences will be severe and global.
”I think that our work shows that early warning signals must be taken very seriously and call for creative and comprehensive use of many observational and model-generated data for better understanding our safe operating spaces, or, namely, how far we are from dangerous tipping behaviour”, concludes Lucarini.
- Journal of Physics A, On some aspects of the response to stochastic and deterministic forcings