Package for process-oriented climate modeling

climate, modeling, modelling, model, ebm, radiation, radiative-convective, earth, climate-model, climate-science, educational-software, python
pip install climlab==0.8.2



Documentation Status JOSS DOI pypi Build Status coverage

Python package for process-oriented climate modeling


Brian E. J. Rose
Department of Atmospheric and Environmental Sciences
University at Albany

About climlab

climlab is a flexible engine for process-oriented climate modeling. It is based on a very general concept of a model as a collection of individual, interacting processes. climlab defines a base class called Process, which can contain an arbitrarily complex tree of sub-processes (each also some sub-class of Process). Every climate process (radiative, dynamical, physical, turbulent, convective, chemical, etc.) can be simulated as a stand-alone process model given appropriate input, or as a sub-process of a more complex model. New classes of model can easily be defined and run interactively by putting together an appropriate collection of sub-processes.

Currently, climlab has out-of-the-box support and documented examples for

  • Radiative and radiative-convective column models, with various radiation schemes:
    • RRTMG (a widely used radiative transfer code)
    • CAM3 (from the NCAR GCM)
    • Grey Gas
    • Simplified band-averaged models (4 bands each in longwave and shortwave)
  • Convection schemes:
    • Emanuel moist convection scheme
    • Hard convective adjustment (to constant lapse rate or to moist adiabat)
  • 1D Advection-Diffusion solvers
  • Moist and dry Energy Balance Models
  • Flexible insolation including: - Seasonal and annual-mean models - Arbitrary orbital parameters
  • Boundary layer scheme including sensible and latent heat fluxes
  • Arbitrary combinations of the above, for example:
    • 2D latitude-pressure models with radiation, horizontally-varying meridional diffusion, and fixed relative humidity


Installing pre-built binaries with conda (Mac OSX, Linux, and Windows)

By far the simplest and recommended way to install climlab is using conda (which is the wonderful package manager that comes with Anaconda Python).

You can install climlab and all its dependencies with:

conda install -c conda-forge climlab

Or (recommended) add conda-forge to your conda channels with:

conda config --add channels conda-forge

and then simply do:

conda install climlab

Binaries are available for OSX, Linux, and Windows.

Installing from source

Consult the documentation for detailed instructions.



These are handled automatically if you install with conda.


  • Python (currently testing on versions 3.8, 3.9, 3.10, 3.11)
  • numpy
  • scipy
  • future
  • pooch (for remote data access and caching)
  • xarray (for data handling)

climlab will still run on Python 2.7 on some systems but we are no longer supporting this

Recommended for full functionality

  • numba >=0.43.1 (used for acceleration of some components)

Note that there is a bug in previous numba versions that caused a hanging condition in climlab under Python 3.

Documentation and Examples

Full user manual is available here.

A rich and up-to-date collection of example usage can be found in Brian Rose's online textbook The Climate Laboratory.

Source notebooks for the tutorials in the docs can be found in the climlab/docs/source/courseware/ directory of the source repo.

These are self-describing, and should run out-of-the-box once the package is installed, e.g:

jupyter notebook Insolation.ipynb

Release history

Version 0.8.2 (released November 2023)

New feature: process class climlab.radiation.InstantInsolation() which correctly interprets longitude, respects local solar time and calculates hour angle. A utility function is also available, with usage mirroring the existing Thanks to @HenryDane for this contribution!

This release also includes numerous bug fixes, updates for Python 3.11, and improvements to documentation and CI builds.

Version 0.8.1 (released May 2022)

A major refactor of the internals: all the Fortran code has been moved into external companion packages climlab-rrtmg, climlab-cam3-radiation, and climlab-emanuel-convection. Climlab is now (once again!) a pure Python package. Builds of these helper packages are available through conda-forge and will be automatically installed as dependencies by conda / mamba.

The climlab source repo also moved to

There should be no breaking changes to the user-facing API.

The major motivation for this change was to (vastly) simplify the development and testing of new-and-improved climlab internals (coming soon).

Version 0.7.13 (released February 2022)

Maintenance release to support Python 3.10.

The attrdict package by Brendan Curran-Johnson has been removed from the dependencies since it is broken on Python 3.10 and no longer under development. A modified version of the MIT-licensed attrdict source is now bundled internally with climlab. There are no changes to climlab's public API.

Version 0.7.12 (released May 2021)
New feature: spectral output from RRTMG (accompanied by a new tutorial)
Version 0.7.11 (released May 2021)
Improvements to data file download and caching (outsourcing this to pooch)
Version 0.7.10 (released April 2021)
Improvements to docs and build.
Version 0.7.9 (released December 2020)
Bug fixes and doc improvements.
Version 0.7.8 (released December 2020)
Bug fixes.
Version 0.7.7 (released October 2020)
Bug fixes.
Version 0.7.6 (released January 2020)
Bug fixes, Python 3.8 compatibility, improvements to build and docs.
Version 0.7.5 (released July 2019)
Bug fixes and improvements to continuous integration
Version 0.7.4 (released June 2019)
New flexible solver for 1D advection-diffusion processes on non-uniform grids, along with some bug fixes.
Version 0.7.3 (released April 2019)
Bug fix and changes to continuous integration for Python 2.7 compatibility
Version 0.7.2 (released April 2019)

Improvements to surface flux processes, a new data management strategy, and improved documentation.

  • climlab.surface.LatentHeatFlux and climlab.surface.SensibleHeatFlux are now documented, more consistent with the climlab API, and have new optional resistance parameters to reduce the fluxes (e.g. for modeling stomatal resistance)
  • climlab.surface.LatentHeatFlux now produces the diagnostic evaporation in kg/m2/s. climlab.convection.EmanuelConvection produces precipitation in the same units.
  • The previous PRECIP diagnostic (mm/day) in climlab.convection.EmanuelConvection is removed. This is a BREAKING CHANGE.
  • Data files have been removed from the climlab source repository. All data is now accessible remotely. climlab will attempt to download and cache data files upon first use.
  • climlab.convection.ConvectiveAdjustement is now accelerated with numba if it is available (optional)
Version 0.7.1 (released January 2019)

Deeper xarray integration, include one breaking change to, Python 3.7 compatibility, and minor enhancements.

  • Removed climlab.utils.attr_dict.AttrDict and replaced with AttrDict package (a new dependency)
  • Added xarray input and output capabilities for
  • and now return xarray.Dataset objects containing the orbital data.
  • The lookup_parameter() method was removed in favor of using built-in xarray interpolation.
  • New class climlab.process.ExternalForcing() for arbitrary externally defined tendencies for state variables.
  • New input option ozone_file=None for radiation components, sets ozone to zero.
  • Tested on Python 3.7. Builds will be available through conda-forge.
Version 0.7.0 (released July 2018)

New functionality, improved documentation, and a few breaking changes to the API.

Major new functionality includes convective adjustment to the moist adiabat and moist EBMs with diffusion on moist static energy gradients.


  • climlab.convection.ConvectiveAdjustement now allows non-constant critical lapse rates, stored in input parameter adj_lapse_rate.
    • New switches to implement automatic adjustment to dry and moist adiabats (pseudoadiabat)
  • climlab.EBM() and its daughter classes are significantly reorganized to better respect CLIMLAB principles:
    • Essentially all the computations are done by subprocesses
    • SW radiation is now handled by climlab.radiation.SimpleAbsorbedShortwave class
    • Diffusion and its diagnostics now handled by climlab.dynamics.MeridionalHeatDiffusion class.
    • Diffusivity can be altered at any time by the user, e.g. during timestepping
    • Diffusivity input value K in class climlab.dynamics.MeridionalDiffusion is now specified in physical units of m2/s instead of (1/s). This is consistent with its parent class climlab.dynamics.Diffusion.
  • A new class climlab.dynamics.MeridionalMoistDiffusion for the moist EBM (diffusion down moist static energy gradient)
  • Tests that require compiled code are now marked with pytest.mark.compiled for easy exclusion during local development

Under-the-hood changes include

  • Internal changes to the timestepping; the compute() method of every subprocess is now called explicitly.
  • compute() now always returns tendency dictionaries
Version 0.6.5 (released April 2018)
Some improved documentation, associated with publication of a meta-description paper in JOSS.
Version 0.6.4 (released February 2018)
Some bug fixes and a new climlab.couple() method to simplify creating complete models from components.
Version 0.6.3 (released February 2018)
Under-the-hood improvements to the Fortran builds which enable successful builds on a wider variety of platforms (incluing Windows/Python3).
Version 0.6.2 (released February 2018)
Introduces the Emanuel moist convection scheme, support for asynchonous coupling, and internal optimzations.
Version 0.6.1 (released January 2018)
Provides basic integration with xarray (convenience methods for converting climlab objects into xarray.DataArray and xarray.Dataset objects)
Version 0.6.0 (released December 2017)
Provides full Python 3 compatibility, updated documentation, and minor enhancements and bug fixes.
Version 0.5.5 (released early April 2017)
Finally provides easy binary distribution with conda
Version 0.5.2 (released late March 2017)
Many under-the-hood improvements to the build procedure, which should make it much easier to get climlab installed on user machines. Binary distribution with conda is coming soon!
Version 0.5 (released March 2017)
Bug fixes and full functionality for the RRTMG radiation module, an improved common API for all radiation modules, and better documentation.
Version 0.4.2 (released January 2017)
Introduces the RRTMG radiation scheme, a much-improved build process for the Fortran extension, and numerous enhancements and simplifications to the API.
Version 0.4 (released October 2016)
Includes comprehensive documentation, an automated test suite, support for latitude-longitude grids, and numerous small enhancements and bug fixes.
Version 0.3 (released February 2016)
Includes many internal changes and some backwards-incompatible changes (hopefully simplifications) to the public API. It also includes the CAM3 radiation module.
Version 0.2 (released January 2015)

The package and its API was completely redesigned around a truly object-oriented modeling framework in January 2015.

It was used extensively for a graduate-level climate modeling course in Spring 2015:

Many more examples are found in the online lecture notes for that course:

Version 0.1

The first versions of the code and notebooks were originally developed in winter / spring 2014 in support of an undergraduate course at the University at Albany.

See the original course webpage at

The documentation was first created by Moritz Kreuzer (Potsdam Institut for Climate Impact Research) as part of a thesis project in Spring 2016.

Contact and Bug Reports

Users are strongly encouraged to submit bug reports and feature requests on github at


This code is freely available under the MIT license. See the accompanying LICENSE file.


Development of climlab is partially supported by the National Science Foundation under award AGS-1455071 to Brian Rose.

Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.