Hybrid physics-ML & parameterization
Data-driven multiscale modeling for correcting dynamical systems
In one sentence. Multiscale networks that pass information fine-to-coarse and coarse-to-fine correct chaotic models, demonstrated on a climate subgrid parameterization task for unresolved fine-scale dynamics.
Abstract
We propose a multiscale approach for predicting quantities in dynamical systems which is explicitly structured to extract information in both fine-to-coarse and coarse-to-fine directions. We envision this method being generally applicable to problems with significant self-similarity or in which the prediction task is challenging and where stability of a learned model's impact on the target dynamical system is important. We evaluate our approach on a climate subgrid parameterization task in which our multiscale networks correct chaotic underlying models to reflect the contributions of unresolved, fine-scale dynamics.
Labels
- Topic
- Hybrid physics-ML & parameterization
- Type
- New model or method
- Method
- Other
- Data
- Other
- Status
- Research
- Identifiers
- arXiv:2303.17496 DOI:10.1088/2632-2153/ae1a36
- Added
- 2026-10-22
Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.