_version_ 1866910045647667200
author Lin, Jerry
Hu, Zeyuan
Beucler, Tom
Frields, Katherine
Christensen, Hannah
Hannah, Walter
Heuer, Helge
Ukkonnen, Peter
Mansfield, Laura A.
Zheng, Tian
Peng, Liran
Gupta, Ritwik
Gentine, Pierre
Al-Naher, Yusef
Duan, Mingjiang
Hattori, Kyo
Ji, Weiliang
Li, Chunhan
Matsuda, Kippei
Murakami, Naoki
Ron, Shlomo
Serlin, Marec
Song, Hongjian
Tanabe, Yuma
Yamamoto, Daisuke
Zhou, Jianyao
Pritchard, Mike
author_facet Lin, Jerry
Hu, Zeyuan
Beucler, Tom
Frields, Katherine
Christensen, Hannah
Hannah, Walter
Heuer, Helge
Ukkonnen, Peter
Mansfield, Laura A.
Zheng, Tian
Peng, Liran
Gupta, Ritwik
Gentine, Pierre
Al-Naher, Yusef
Duan, Mingjiang
Hattori, Kyo
Ji, Weiliang
Li, Chunhan
Matsuda, Kippei
Murakami, Naoki
Ron, Shlomo
Serlin, Marec
Song, Hongjian
Tanabe, Yuma
Yamamoto, Daisuke
Zhou, Jianyao
Pritchard, Mike
contents Subgrid machine-learning (ML) parameterizations have the potential to introduce a new generation of climate models that incorporate the effects of higher-resolution physics without incurring the prohibitive computational cost associated with more explicit physics-based simulations. However, important issues, ranging from online instability to inconsistent online performance, have limited their operational use for long-term climate projections. To more rapidly drive progress in solving these issues, domain scientists and machine learning researchers opened up the offline aspect of this problem to the broader machine learning and data science community with the release of ClimSim, a NeurIPS Datasets and Benchmarks publication, and an associated Kaggle competition. This paper reports on the downstream results of the Kaggle competition by coupling emulators inspired by the winning teams' architectures to an interactive climate model (including full cloud microphysics, a regime historically prone to online instability) and systematically evaluating their online performance. Our results demonstrate that online stability in the low-resolution, real-geography setting is reproducible across multiple diverse architectures, which we consider a key milestone. All tested architectures exhibit strikingly similar offline and online biases, though their responses to architecture-agnostic design choices (e.g., expanding the list of input variables) can differ significantly. Multiple Kaggle-inspired architectures achieve state-of-the-art (SOTA) results on certain metrics such as zonal mean bias patterns and global RMSE, indicating that crowdsourcing the essence of the offline problem is one path to improving online performance in hybrid physics-AI climate simulation.
format Preprint
id arxiv_https___arxiv_org_abs_2511_20963
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Crowdsourcing the Frontier: Advancing Hybrid Physics-ML Climate Simulation via a $50,000 Kaggle Competition
Lin, Jerry
Hu, Zeyuan
Beucler, Tom
Frields, Katherine
Christensen, Hannah
Hannah, Walter
Heuer, Helge
Ukkonnen, Peter
Mansfield, Laura A.
Zheng, Tian
Peng, Liran
Gupta, Ritwik
Gentine, Pierre
Al-Naher, Yusef
Duan, Mingjiang
Hattori, Kyo
Ji, Weiliang
Li, Chunhan
Matsuda, Kippei
Murakami, Naoki
Ron, Shlomo
Serlin, Marec
Song, Hongjian
Tanabe, Yuma
Yamamoto, Daisuke
Zhou, Jianyao
Pritchard, Mike
Atmospheric and Oceanic Physics
Machine Learning
Subgrid machine-learning (ML) parameterizations have the potential to introduce a new generation of climate models that incorporate the effects of higher-resolution physics without incurring the prohibitive computational cost associated with more explicit physics-based simulations. However, important issues, ranging from online instability to inconsistent online performance, have limited their operational use for long-term climate projections. To more rapidly drive progress in solving these issues, domain scientists and machine learning researchers opened up the offline aspect of this problem to the broader machine learning and data science community with the release of ClimSim, a NeurIPS Datasets and Benchmarks publication, and an associated Kaggle competition. This paper reports on the downstream results of the Kaggle competition by coupling emulators inspired by the winning teams' architectures to an interactive climate model (including full cloud microphysics, a regime historically prone to online instability) and systematically evaluating their online performance. Our results demonstrate that online stability in the low-resolution, real-geography setting is reproducible across multiple diverse architectures, which we consider a key milestone. All tested architectures exhibit strikingly similar offline and online biases, though their responses to architecture-agnostic design choices (e.g., expanding the list of input variables) can differ significantly. Multiple Kaggle-inspired architectures achieve state-of-the-art (SOTA) results on certain metrics such as zonal mean bias patterns and global RMSE, indicating that crowdsourcing the essence of the offline problem is one path to improving online performance in hybrid physics-AI climate simulation.
title Crowdsourcing the Frontier: Advancing Hybrid Physics-ML Climate Simulation via a $50,000 Kaggle Competition
topic Atmospheric and Oceanic Physics
Machine Learning
url https://arxiv.org/abs/2511.20963