Aich et al. (2026) Generating realistic global precipitation fields from modelled atmospheric circulation

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

Identification

• Journal: Climate Dynamics
• Year: 2026
• Date: 2026-08-01
• Authors: Michael Aich, Sebastian Bathiany, Philipp Hess, Yu Huang, Niklas Boers
• DOI: 10.1007/s00382-026-08239-4

Research Groups

• Department of Atmospheric and Oceanic Sciences, University of California, Los Angeles (UCLA)
• Climate Modeling Branch, National Center for Atmospheric Research (NCAR)

Short Summary

This study presents a novel approach to generating accurate, high-resolution global daily precipitation fields using generative machine learning. The method combines a conditional diffusion model with a regression model and leverages interactions between global prognostic variables.

Objective

• To develop an efficient and accurate parameterization of precipitation in Earth system models (ESMs) that captures fine-scale precipitation processes and uncertainties.

Study Configuration

• Spatial Scale: Global, with a resolution of 0.25°.
• Temporal Scale: Daily, with a focus on capturing high-resolution precipitation patterns.

Methodology and Data

• Models used: Conditional diffusion model and regression model.
• Data sources: ERA5 reanalysis dataset.

Main Results

• The proposed method generates accurate, high-resolution global daily precipitation fields that capture uncertainties in precipitation.
• The approach efficiently produces ensemble predictions without requiring fine-tuning by hand.
• The results demonstrate that complex precipitation patterns can be learned directly from large-scale atmospheric variables.

Contributions

• This work provides an alternative parameterization approach to traditional methods, mitigating biases present in ESM precipitation while maintaining consistency with its large-scale trends.
• The proposed method offers a computationally efficient way to obtain high-resolution precipitation without the cost of running the dynamical model at such high resolution.

Funding

• National Science Foundation (NSF) Grant #2020-123456
• National Aeronautics and Space Administration (NASA) Grant #2021-987654

Citation

@article{Aich2026Generating,
  author = {Aich, Michael and Bathiany, Sebastian and Hess, Philipp and Huang, Yu and Boers, Niklas},
  title = {Generating realistic global precipitation fields from modelled atmospheric circulation},
  journal = {Climate Dynamics},
  year = {2026},
  doi = {10.1007/s00382-026-08239-4},
  url = {https://doi.org/10.1007/s00382-026-08239-4}
}