CDS Team Members Lead AI Initiative to Enhance Climate Change Projections
CDS affiliated professor of atmosphere/ocean science and mathematics Laure Zanna is leading a project supported by Schmidt Futures — a philanthropic initiative founded by Eric and Wendy Schmidt that bets early on exceptional people making the world better, particularly through innovative breakthroughs in science and technology. The project’s objective is to enhance climate-change projections by improving climate simulations using artificial intelligence (AI). Additionally, Carlos Fernanda-Granda, assistant professor of mathematics and data science at NYU’s Courant Institute of Mathematical Science and CDS and Joan Bruna, associate professor at NYU’s Courant Institute, in the Department of Computer Science, Department of Mathematics (affiliated), and CDS will be leading the AI innovation for climate component of the project.
“Despite drastic improvements in climate model development, current simulations have difficulty capturing the interactions among different processes in the atmosphere, oceans, and ice and how they affect the Earth’s climate; this can hinder projections of temperature, rainfall, and sea level,” describes Zanna, “AI and machine-learning tools excel at extracting complex information from data and will help bolster the accuracy of our climate simulations and predictions to better inform the work of policymakers and scientists.”
About the Team’s Background
Laure Zanna’s research focuses on the dynamics of the climate system. The main objective of her work is to study the influence of the ocean on both local and global scales through the analysis of observations and a hierarchy of numerical simulations.
Joan Bruna’s research interests surrounds Machine Learning, Signal Processing, and High-Dimensional Statistics. The past few years, Bruna has been working on Deep Learning, studying some of its theoretical properties. He’s interested in generalisations of CNNs to more general geometries and their applications to Physics, Chemistry and Computational Complexity.
Carlos Fernandez-Granda’s research focuses on the design and analysis of data-science methodology. He’s interested in theory of inverse problems, machine learning, and data-driven medicine.
The Problem to be Solved
Scientists rely on computer simulations (or climate models) to describe the evolution of atmosphere, ocean, and ice systems due to their complexity. These climate models separate the climate system into a series of grid boxes (or grid cells) to imitate how the ocean, atmosphere, and ice are changing and interact with each other. The issue is that the number of grid boxes selected is limited by computer power. Currently, climate models for multi-decade projections use grid box sizes measuring approximately 50 km to 100 km (roughly 30 to 60 miles). As a result, processes that occur on scales that are smaller than the grid cell (clouds, turbulence, and ocean mixing) are in turn not well captured.
The initiative essentially will use machine learning to more holistically capture physical ocean, ice, and atmosphere processes as a means to reduce the imprecision of existing models. Ultimately, machine learning will lead the development of interpretable, physics-guided representations of these complex processes directly from data for use in global climate simulations.
“We will develop machine-learning methodology to automatically infer high-resolution dynamics from low-resolution measurements,” explains Fernandez-Granda. “The challenge is to design the methodology so that it is consistent with the underlying physics, and is as interpretable as possible.”
For more information about Schmidt Futures, please visit the Schmidt Futures website.
By Ashley C. McDonald
