Zheng et al. (2026) Climate‐Adaptive and Cascade‐Constrained Machine Learning Prediction for Sea Surface Height Under Greenhouse Warming

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Journal of Advances in Modeling Earth Systems
• Year: 2026
• Date: 2026-07-01
• Authors: Ru Chen, Xin Su, Julian Mak, Gang Huang
• DOI: 10.1029/2025ms005717

Research Groups

• Ocean Modeling Group (OMG), University of Tokyo
• Climate Science Department, National Institute for Environmental Studies (NIES)

Short Summary

This study evaluates the climate adaptability of machine learning (ML) models in predicting sea surface height (SSH) using kinetic energy (KE) cascade as a physical constraint. The results show improved model performance under greenhouse warming conditions.

Objective

• Investigate the impact of incorporating KE cascade as a physical constraint on ML-based ocean state prediction

Study Configuration

• Spatial Scale: Regional, focusing on the Kuroshio Extension and Gulf Stream regions
• Temporal Scale: Present climate and future (greenhouse warming) scenarios

Methodology and Data

• Models used: Two SSH prediction models: one with KE cascade constraint and one without
• Data sources: Satellite observations, reanalysis datasets

Main Results

• Incorporating KE cascade as a physical constraint improves model performance by 19.5% in the present climate and 19.6% under greenhouse warming conditions
• Reduced eddy kinetic energy errors in both models under greenhouse warming conditions
• Improved cross-scale transfer of KE compared to traditional KE spectrum constraint

Contributions

• First application of KE cascade as a physical constraint for ML-based ocean state prediction
• Demonstrates robust adaptability across climates, offering guidance for global ML model development

Funding

• This research was supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI Grant Number 19H04252 and the National Institute for Environmental Studies (NIES) Research Program.

Citation

@article{Zheng2026ClimateAdaptive,
  author = {Zheng, Tianmu and Chen, Ru and Su, Xin and Mak, Julian and Huang, Gang and Yan, Bingzheng},
  title = {Climate‐Adaptive and Cascade‐Constrained Machine Learning Prediction for Sea Surface Height Under Greenhouse Warming},
  journal = {Journal of Advances in Modeling Earth Systems},
  year = {2026},
  doi = {10.1029/2025ms005717},
  url = {https://doi.org/10.1029/2025ms005717}
}