v0.9.4 · BSD licensed units-aware

Build Earth-system models in Python.

climt is a Climate Modelling and Diagnostics Toolkit — a modular, intuitive way to write numerical models of the climate system. State-of-the-art components, no Fortran required.

Get started → Read the docs
$ pip install climt ⧉
climt jigsaw-globe logo
Why climt

Modelling, made composable

Plug-and-play

Mix and match independent physical components — radiation, convection, surface — into a model. No rewrites between levels of complexity.

Units-aware

Built on sympl and Pint. Every quantity carries its units; xarray DataArrays make model arrays self-describing.

Self-documenting

Descriptive names in the user API mean model scripts read like prose — reproducible and easy to learn.

State of the art

RRTMG radiation, Emanuel convection, correlated-k schemes — research-quality components, accessible to students.

A model in five lines

Create components, allocate a state, and run. The defaults are scientifically meaningful, so your first model does something sensible immediately.

scientifically-sane defaults
●first_model.py
import climt

radiation = climt.GrayLongwaveRadiation()
surface   = climt.SlabSurface()

state = climt.get_default_state([radiation, surface])
tendencies, diagnostics = radiation(state)