fme: Full Model Emulation

fme (“full model emulation”) is a python package for training, running and evaluating climate model emulators, such as the Ai2 Climate Emulator.

Why use fme?

• fme provides a unified interface for training, running and evaluating AI models with a range of architectures (SFNO, GNNs, UNets) and applications (atmosphere modeling, ocean modeling, downscaling).

• Built by climate modelers for climate modelers! We follow similar configuration and evaluation practices as traditional climate models, making fme intuitive to adopt and use.

• Flexible data loading and writing capabilities. fme supports netCDF and zarr, as well as streaming directly from/to cloud object storage. At inference time, reductions such as monthly means can be computed on the fly, saving time and storage space.

Contents:

• Installation
• Quickstart
• Training Config
• Inference Config
• Downscaling Inference
• Evaluator Config
• Coupled Modeling
• Accessing data via Google Cloud Storage
• The Builder Pattern: Modern Namelists
• Modules
• Steps
• API Reference