Incubator Application Questionnaire: wradlib

[Heistermann]

Incubator Application Questionnaire: wradlib

1. Please provide the name and email address of the principal Project Owner.

• Maik Heistermann, maik.heistermann@…
• Kai Muehlbauer, kai.muehlbauer@…

2. Please provide the names and emails of co-project owners (if any).

We have no formal difference between owner or co-owner.

3. Please provide the names, emails and entity affiliation of all official committers

• Maik Heistermann, maik.heistermann@…, University of Potsdam, Germany
• Kai Muehlbauer, kai.muehlbauer@…, University of Bonn, Germany
• Jonathan J. Helmus, jjhelmus@…, Argonne National Laboratory, IL, USA
• Thomas Pfaff, thomaspfaff@…, Blue Yonder, Karlsruhe, Germany

4. Please describe your Project.

Terrestrial weather radars have become an invaluable tool for remote sensing of the atmosphere both for research and applications. The main (though not exclusive) purpose of this technology is to provide precipitation information at high spatiotemporal resolution and large areal coverage.

wradlib is an Open-Source Python library designed to support users in developing processing chains for weather radar data that are custom-tailored to their application context. For this purpose, wradlib addresses a variety of challenges, including the ingestion of common data formats, geo-integration of observations from a spherical coordinate system with the added complexity of earth curvature and atmospheric refractivity, conversion of reflectivity to rainfall intensity, detection and correction of typical error sources and artifacts, merging with other sensors, and visualisation.

The overall point of entry is ​http://wradlib.org. The code is hosted at ​https://github.com/wradlib/wradlib, the documentation is available at ​http://wradlib.org/wradlib-docs. A user mailing list/ forum is maintained at ​https://groups.google.com/forum/?fromgroups#!forum/wradlib-users.

See scientific reference e.g. in ​Heistermann et al. 2013 or ​Heistermann et al. 2015.

5. Why is hosting at OSGeo good for your project?

• Seeking support for enhancing the professional standards of software development;
• Enhancing project visibility within and beyond the OSGeo community;
• Networking and integration with other geospatial software packages, both in terms of developers and user communities - the Python gdal package

already is a major dependency and we would hope to intensify such linkages in order to maximise synergies. A recent example of collaboration of wradlib developers with osgeo gdal developers is the implementation of "GEOS prepared geometry in gdal layer methods" (​https://trac.osgeo.org/gdal/ticket/6323).

6. Type of application does this project represent (client, server, standalone, library, etc.):

Library

7. Please describe any relationships to other open source projects.

The dependencies of wradlib are numpy, scipy, matplotlib, netCDF4, h5py, and gdal. In particular, the gdal package is wradlib's major backbone for geointegration and geoprocessing.

In addition, wradlib maintains strong ties with other Open Source packages for weather radar processing, namely Py-ART and BALTRAD (see e.g. ​http://openradar.github.io, or ​Heistermann et al. 2015).

8. Please describe any relationships with commercial companies or products.

None so far. There are applications of wradlib by users in the commercial sector, though.

9. Which open source license(s) will the source code be released under?

MIT

10. Is there already a beta or official release?

The current release at the time of writing is 0.8.0

11. What is the origin of your project (commercial, experimental, thesis or other higher education, government, or some other source)?

The project is a result of a research cooperation between the universities of Potsdam, Bonn and Stuttgart.

12. Does the project support open standards? Which ones and to what extent? (OGC, w3c, ect.) Has the software been certified to any standard (CITE for example)? If not, is it the intention of the project owners to seek certification at some point?

wradlib does not explicitely support standards. It supports reading various open data models developed by the weather radar community, namely CfRadial (NetCDF format) and ODIM_H5 (hdf5 format). Reading the actual file formats is achieved by using the Python packages netcdf4 and h5py as dependencies.

13. Is the code free of patents, trademarks, and do you control the copyright?

Yes.

14. How many people actively contribute (code, documentation, other?) to the project at this time?

• Code: 3 people
• Research: 5 people

15. How many people have commit access to the source code respository?

2 people (owners, see point 1)

16. Approximately how many users are currently using this project?

96 users are registered to the wradlib mailing list as of the time of filling out this questionnaire. As wradlib is the main radar processing library used at the University of Bonn and Potsdam, graduate students, PhD students, and associated researchers are using wradlib for their work.

17. What type of users does your project attract (government, commercial, hobby, academic research, etc. )?

Very diverse. There is no formal registration, but use cases have been reported from all around the world from academia, government authorities, private companies, and private users, too. The main user fraction is from academia and research, though.

18. If you do not intend to host any portion of this project using the OSGeo infrastructure, why should you be considered a member project of the OSGeo Foundation?

We are not opposing to move portions of this project to OSGeo infrastructure if the overall benefits exceed the costs. We expect such a discussion to be part of the incubation process. Still, we just recently moved our hosting infrastructure from bitbucket to github and our CI from Codeship to Travis, and we are quite happy with this.

19. Does the project include an automated build and test?

We use Travis for automated testing and deployment of documentation and new releases via PyPI.

20. What language(s) are used in this project? (C/Java/perl/etc)

Python

21. What is the dominant written language (i.e. English, French, Spanish, German, etc) of the core developers?

English

22. What is the (estimated) size of a full release of this project? How many users do you expect to download the project when it is released?

Right now, the size of the 0.8.0 release is about 24 MB. At ​PyPI, wradlib has achieved 756 downloads over the last month, and 429 downloads over the last week (at the time of filling out this questionnaire).

[jive]

I note this project has not yet attracted a mentor, please reach out to us on the incubation email list - we have a new OSGeo Community program that may be more appropriate place to get started.

[kmuehlbauer]

@jive I'm just adding the reference here for the quite aged thread at the ​incubator list. We can post an update on wradlib there shortly, if needed.