paper
2018
Hybrid physics-ML & parameterization
Climate emulation New model or method
OtherClimate projections / centennialColumbia University (LEAP, Lamont-Doherty) Other
2018-03-16
030
Mar 2018
Pierre Gentine, Mike Pritchard, Stephan Rasp et al.
preprint
Columbia University (LEAP, Lamont-Doherty)
CBRAIN, a neural network trained on superparameterized aquaplanet data, replaces convection parameterization and reproduces its features with reduced stochasticity.
Hybrid physics-ML & parameterization
Could machine learning break the convection parametrization deadlock?
In one sentence. CBRAIN, a neural network trained on superparameterized aquaplanet data, replaces convection parameterization and reproduces its features with reduced stochasticity.
Labels
- Topic
- Hybrid physics-ML & parameterization, Climate emulation
- Type
- New model or method
- Method
- Other
- Time scale
- Climate projections / centennial
- Domain
- Global
- Data
- km-scale or high-res simulation
- Status
- Research
- Identifiers
- DOI:10.31223/osf.io/hv95e OpenAlex:W4239236185
- Added
- 2026-10-22
Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.