MetPy is an Open Source project aimed at providing a Pythonic library for meteorological data analysis that meshes well with the rest of the scientific Python ecosystem. The project heavily leverages the work already done by the Numpy, Scipy, and Matplotlib projects, and adds on top functionality specific to meteorology: plotting (e.g. Skew-Ts), calculations, and reading files (e.g. WSR-88D NIDS files).
The MetPy project began at the University of Oklahoma when then-graduate students Ryan May and Sean Arms (now Unidata Program Center software developers), together with fellow graduate students Patrick Marsh and Eric Bruning, thought it would be a good way to avoid working on their doctoral theses. After years of dormancy due to said theses, work on MetPy has recently resumed.
While MetPy is not an official project of the Unidata Program Center, May and Arms are both heavily involved. Participation by Unidata community members — anything from trying the code and providing feedback to contributing code — is heartily encouraged.