Hybrid physics-ML & parameterization
Towards Physically-consistent, Data-driven Models of Convection
In one sentence. Enforces physical constraints in neural convection emulators via loss or architecture, and rescales data so models generalize to unseen climates.
Abstract
Data-driven algorithms, in particular neural networks, can emulate the effect of sub-grid scale processes in coarse-resolution climate models if trained on high-resolution climate simulations. However, they may violate key physical constraints and lack the ability to generalize outside of their training set. Here, we show that physical constraints can be enforced in neural networks, either approximately by adapting the loss function or to within machine precision by adapting the architecture. As these physical constraints are insufficient to guarantee generalizability, we additionally propose to physically rescale the training and validation data to improve the ability of neural networks to generalize to unseen climates.
Labels
- Topic
- Hybrid physics-ML & parameterization, Explainability, uncertainty & physics
- Type
- New model or method
- Method
- Physics-informed / physics-constrained
- Time scale
- Climate projections / centennial
- Domain
- Global
- Data
- km-scale or high-res simulation
- Status
- Research
- Identifiers
- arXiv:2002.08525
- Added
- 2026-10-22
Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.