The Python-ARM Radar Toolkit. A data model driven interactive toolkit for working with weather radar data.
Python Jupyter Notebook C Other
Latest commit 4481ce3 Mar 4, 2017 @jjhelmus jjhelmus committed on GitHub Merge pull request #640 from jjhelmus/no_jjhelmus_ci
TST: Do not use jjhelmus channel for Travis CI test environment


Travis AppVeyor

The Python ARM Radar Toolkit (Py-ART)

The Python ARM Radar Toolkit, Py-ART, is an open source Python module containing a growing collection of weather radar algorithms and utilities build on top of the Scientific Python stack and distributed under the 3-Clause BSD license. Py-ART is used by the Atmospheric Radiation Measurement (ARM) Climate Research Facility for working with data from a number of precipitation and cloud radars, but has been designed so that it can be used by others in the radar and atmospheric communities to examine, processes, and analyze data from many types of weather radars.

Important Links


If you use the Python ARM Radar Toolkit (Py-ART) to prepare a publication please cite:

Helmus, J.J. & Collis, S.M., (2016). The Python ARM Radar Toolkit (Py-ART), a Library for Working with Weather Radar Data in the Python Programming Language. Journal of Open Research Software. 4(1), p.e25. DOI:

Py-ART implements many published scientific methods which should also be cited if you make use of them. Refer to the References section in the documentation of the functions used for information on these citations.


The easiest method for installing Py-ART is to use the conda packages from the latest release. To do this you must download and install Anaconda or Miniconda. Then use the following command in a terminal or command prompt to install the latest version of Py-ART:

conda install -c conda-forge arm_pyart

To update an older version of Py-ART to the latest release use:

conda update -c conda-forge arm_pyart

If you do not wish to use Anaconda or Miniconda as a Python environment or want to use the latest, unreleased version of Py-ART see the section below on Installing from source.


The configuration file in Py-ART specifies the default metadata, field names, colormaps and plot limits. A custom configuration can be loaded automatically be setting the environmental variable PYART_CONFIG to point to a custom configuration file. For additional details on this process see the documentation on the pyart.load_config function.

Extensions and related software

A number of projects are available which extend the functionality of Py-ART. These include:

  • ARTView : Interactive radar viewing browser.
  • PyTDA : Python Turbulence Detection Algorithm.
  • SingleDop : Single Doppler Retrieval Toolkit.
  • DualPol : Python Interface to Dual-Pol Radar Algorithms.
  • PyBlock: Python Polarimetric Radar Beam Blockage Calculation

Other related open source software for working with weather radar data:

  • wradlib : An open source library for weather radar data processing.
  • BALTRAD : Community-based weather radar networking.
  • MMM-Py : Marshall MRMS Mosaic Python Toolkit.
  • CSU_RadarTools : Colorado State University Radar Tools.
  • TRMM RSL : TRMM Radar Software Library.
  • RadX: Radx C++ Software Package for Radial Radar Data.


Py-ART is tested to work under Python 2.7, 3.4, 3.5, and 3.6.

The required dependencies to install Py-ART in addition to Python are:

As well as a working C/C++ compiler. A Fortran compiler is required for some optional modules. An easy method to install these dependencies is by using a Scientific Python distributions. Anaconda will install all of the above packages by default on Windows, Linux and Mac computers and is provided free of charge by Continuum Analytics.

Optional Dependences

The above Python modules are require before installing Py-ART, additional functionality is available of the following modules are installed.

  • TRMM Radar Software Library (RSL). If installed Py-ART will be able to read in radar data in a number of additional formats (Lassen, McGill, Universal Format, and RADTEC) and perform automatic dealiasing of Doppler velocities. RSL should be install prior to installing Py-ART. The environmental variable RSL_PATH should point to the location where RSL was installed if RSL was not installed in the default location (/usr/local/trmm).
  • In order to read files which are stored in HDF5 files the h5py package and related libraries must be installed.
  • A linear programming solver and Python wrapper to use the LP phase processing method. CyLP is recommended as it gives the fastest results, but PyGLPK and CVXOPT are also supported. The underlying LP solvers CBC or GLPK will also be required depending on which wrapper is used.
  • Basemap. If installed the ability to plot grids on geographic maps is available.
  • wradlib. Needed to calculate the texture of a differential phase field.
  • nose. Required to run the Py-ART unit tests.
  • gdal. Required to output GeoTIFFs from Grid objects.

Installing from source

Installing Py-ART from source is the only way to get the latest updates and enhancement to the software that have not yet made it into a release. The latest source code for Py-ART can be obtained from the GitHub repository, Either download and unpack the zip file of the source code or use git to checkout the repository:

git clone

To install in your home directory, use:

python install --user

To install for all users on Unix/Linux:

python build
sudo python install


Py-ART is an open source, community software project. Contributions to the package are welcomed from all users.


The latest source code can be obtained with the command:

git clone

If you are planning on making changes that you would like included in Py-ART, forking the repository is highly recommended.


We welcome contributions for all used of Py-ART provided the code can be distributed under the BSD 3-clause license. A copy of this license is available in the LICENSE.txt file in this directory.


After installation, you can launch the test suite from outside the source directory (you will need to have nosetests installed):

$ nosetests --exe pyart

In-place installs can be tested using the nosetest command from within the source directory.