Tackling Climate Change with Machine Learning

Published in 2019, this paper surveys the landscape of machine learning applications relevant to climate change. It systematically covers how ML can contribute to climate mitigation (reducing emissions) and adaptation (responding to climate impacts) across multiple domains including energy systems, agriculture, forests, infrastructure, and climate science itself.

The authors identify high-impact opportunities where ML can accelerate progress, such as optimizing renewable energy grids, improving crop yields with precision agriculture, and enhancing climate modeling. They also discuss barriers to deployment, including data availability, computational resources, and the need for domain expertise.

The work has become influential in establishing ML for climate as a distinct research area and has motivated substantial follow-up work in both academia and industry. It serves as a reference point for researchers and practitioners seeking to understand where machine learning can make the most difference in addressing climate challenges.