pycmor

• Simplifies CMIP6/7 compliance — converts Earth system model output into
  CMOR-standard NetCDF without requiring direct use of the low-level CMOR library.
• Model-agnostic pipeline — plugin-based architecture works with any ESM;
  ships with support for FESOM2, OpenIFS, and PISM out of the box.
• HPC-native — scales across SLURM clusters via dask.distributed, handling
  large multi-variable, multi-frequency datasets in parallel.
• CF-conventions and WCRP Controlled Vocabularies built in — automatically
  applies correct metadata, coordinates, bounds, and calendar handling.
• CMIP7-ready — active development targets full CMIP7 data request and CV
  compliance, tracking the emerging WCRP standard.
• Extensible by design — researchers and modelling groups can add their own
  pipeline steps, variable mappings, and model plugins without modifying core code.

What is pycmor?

pycmor (installed as py-cmor) is a Python package developed at the
Alfred Wegener Institute (AWI) Helmholtz Centre for Polar and Marine Research.
It provides a declarative, rule-based pipeline for reformatting climate model
output into CMOR-compliant NetCDF files required for
submission to CMIP6 and CMIP7.

The package is structured around a YAML-driven configuration system: users
describe which model variables map to which CMIP variables, and pycmor handles
coordinate generation, time bounds, calendar conversion, chunking, and metadata
injection automatically.

Getting started

Install from PyPI:

pip install py-cmor

Then run the CMORisation pipeline with a config file:

pymor process config.yaml

Full documentation is available at pymor.readthedocs.io.
Source code and issue tracker: github.com/esm-tools/pycmor.