Machine Learning for the Physics of Climate

Annalisa Bracco,Julien Brajard,Henk A. Dijkstra,Pedram Hassanzadeh,Christian Lessig,Claire Monteleoni

2024-08-19

Abstract:An exponential growth in computing power, which has brought more sophisticated and higher resolution simulations of the climate system, and an exponential increase in observations since the first weather satellite was put in orbit, are revolutionizing climate science. Big data and associated algorithms, coalesced under the field of Machine Learning (ML), offer the opportunity to study the physics of the climate system in ways, and with an amount of detail, infeasible few years ago. The inference provided by ML has allowed to ask causal questions and improve prediction skills beyond classical barriers. Furthermore, when paired with modeling experiments or robust research in model parameterizations, ML is accelerating computations, increasing accuracy and allowing for generating very large ensembles at a fraction of the cost. In light of the urgency imposed by climate change and the rapidly growing role of ML, we review its broader accomplishments in climate physics. Decades long standing problems in observational data reconstruction, representation of sub-grid scale phenomena and climate (and weather) prediction are being tackled with new and justified optimism. Ultimately, this review aims at providing a perspective on the benefits and major challenges of exploiting ML in studying complex systems.

Atmospheric and Oceanic Physics,Computational Physics

What problem does this paper attempt to address?

The paper attempts to address several long-standing challenges in climate change prediction, including the reconstruction of observational data, the representation of subgrid-scale phenomena, and climate (and weather) forecasting. Specifically:

1. **Reconstruction of Observational Data**: How to use machine learning methods to extend or better interpret observational data, particularly in terms of temporal, spatial, and observational variable extensions.
2. **Subgrid-Scale Parameterization**: How to improve the parameterization of small-scale processes (such as turbulent motion) through machine learning methods to enhance the accuracy and efficiency of climate models.
3. **Climate and Weather Forecasting**: How to use machine learning methods to overcome classical prediction barriers, improve short-term weather forecasts and long-term climate predictions, while reducing the demand for computational resources.

These issues are critical problems that need to be addressed in climate change research. The application of machine learning in these areas has already achieved some significant breakthroughs and is expected to further advance climate science.