Climate-Invariant Machine Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Beucler, Tom, Gentine, Pierre, Yuval, Janni, Gupta, Ankitesh, Peng, Liran, Lin, Jerry, Yu, Sungduk, Rasp, Stephan, Ahmed, Fiaz, O'Gorman, Paul A., Neelin, J. David, Lutsko, Nicholas J., Pritchard, Michael
Format: Preprint
Published: 2021
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910300037447680
author Beucler, Tom
Gentine, Pierre
Yuval, Janni
Gupta, Ankitesh
Peng, Liran
Lin, Jerry
Yu, Sungduk
Rasp, Stephan
Ahmed, Fiaz
O'Gorman, Paul A.
Neelin, J. David
Lutsko, Nicholas J.
Pritchard, Michael
author_facet Beucler, Tom
Gentine, Pierre
Yuval, Janni
Gupta, Ankitesh
Peng, Liran
Lin, Jerry
Yu, Sungduk
Rasp, Stephan
Ahmed, Fiaz
O'Gorman, Paul A.
Neelin, J. David
Lutsko, Nicholas J.
Pritchard, Michael
contents Projecting climate change is a generalization problem: we extrapolate the recent past using physical models across past, present, and future climates. Current climate models require representations of processes that occur at scales smaller than model grid size, which have been the main source of model projection uncertainty. Recent machine learning (ML) algorithms hold promise to improve such process representations, but tend to extrapolate poorly to climate regimes they were not trained on. To get the best of the physical and statistical worlds, we propose a new framework - termed "climate-invariant" ML - incorporating knowledge of climate processes into ML algorithms, and show that it can maintain high offline accuracy across a wide range of climate conditions and configurations in three distinct atmospheric models. Our results suggest that explicitly incorporating physical knowledge into data-driven models of Earth system processes can improve their consistency, data efficiency, and generalizability across climate regimes.
format Preprint
id arxiv_https___arxiv_org_abs_2112_08440
institution arXiv
publishDate 2021
record_format arxiv
spellingShingle Climate-Invariant Machine Learning
Beucler, Tom
Gentine, Pierre
Yuval, Janni
Gupta, Ankitesh
Peng, Liran
Lin, Jerry
Yu, Sungduk
Rasp, Stephan
Ahmed, Fiaz
O'Gorman, Paul A.
Neelin, J. David
Lutsko, Nicholas J.
Pritchard, Michael
Machine Learning
Atmospheric and Oceanic Physics
Computational Physics
Projecting climate change is a generalization problem: we extrapolate the recent past using physical models across past, present, and future climates. Current climate models require representations of processes that occur at scales smaller than model grid size, which have been the main source of model projection uncertainty. Recent machine learning (ML) algorithms hold promise to improve such process representations, but tend to extrapolate poorly to climate regimes they were not trained on. To get the best of the physical and statistical worlds, we propose a new framework - termed "climate-invariant" ML - incorporating knowledge of climate processes into ML algorithms, and show that it can maintain high offline accuracy across a wide range of climate conditions and configurations in three distinct atmospheric models. Our results suggest that explicitly incorporating physical knowledge into data-driven models of Earth system processes can improve their consistency, data efficiency, and generalizability across climate regimes.
title Climate-Invariant Machine Learning
topic Machine Learning
Atmospheric and Oceanic Physics
Computational Physics
url https://arxiv.org/abs/2112.08440