A new study by scientists in the UK, France and Israel has found that AI programs used widely in climate science build an actual understanding of the climate system, meaning we can trust machine learning and further its applications in climate science.
Large, complex climate models are often impractical to work with as they need to run for months on supercomputers. As an alternative, climate scientists often study simplified models.
Generally, two different approaches are used to simplify climate models: a top-down approach where climate experts estimate what impact left out functions will have on the parts kept in the reduced model; and a bottom-up approach, where climate data is fed a machine learning program, which then simulates the climate system.
The two methods turn out comparable results. It is a challenging problem, however, to physically understand data-driven (bottom-up) approaches to fully trust them. Do machine learning programs “understand” that they are dealing with a complex dynamical system, or are they simply good at statistically guessing the right answers?
The new research published in the journal Chaos by the University of Reading, UK, the Weizmann Institute, Israel, and the Ecole Normale Supérieure, France, showed using computer simulations that a machine learning program called Empirical Model Reduction (EMR) in fact knows what it is doing.
The study shows that this computer program reaches comparable results to the top-down reductions of larger models because machine learning constructs its own version of a climate model in its software.
Manuel Santos Gutiérrez, a PhD student at the University of Reading, said, “I think what we do in this investigation is give some sort of physical evidence of why this particular data-driven protocol works. And that to me is quite meaningful, because the method has been in the atmospheric sciences for quite a long time. Yet there was still quite a lot of gaps in the understanding of the methodologies.”
The study indicates that the machine learning method is dynamically and physically sound and produces robust simulations. According to the authors, this should motivate the further use of data-driven methods in climate science as well as other sciences.
The study is part of the European Horizon 2020 TiPES project on tipping points in the Earth system. TiPES is administered from the University of Copenhagen, Denmark.
To read the full study, click here.
