Thanks to generous financial support provided by Unidata in summer 2021, the University of Wisconsin-Milwaukee Atmospheric Science Program purchased one Dell PowerEdge T640 server to enable us to continue our IDD participation and deploy a local The Littlest JupyterHub instance. Since its deployment in August 2021, the server, data we receive through the IDD, and the local JupyterHub instance have supported student training and development in four Atmospheric Science courses and through our Innovative Weather student traineeship program.
We chose The Littlest JupyterHub because it was explicitly designed for our most-frequent use cases: a small use base (up to 50 students), running on a single node (in this case, the PowerEdge T640 server), providing the same instructor-controlled environment for all students and users, with limited required maintenance. The installation process also seemed straightforward compared to that for a full JupyterHub distribution.
Our undergraduate Atmospheric Science major transitioned its programming requirement from FORTRAN to Python effective in Fall 2021. This allowed us to modernize our two-semester Synoptic Meteorology I and II lab curriculum, wherein students gain experience in generating meteorological analyses using Python rather than using instructor- and/or web-generated analyses, and to develop a new lab-based course, Introduction to Climate Science, wherein students use Python resources to interrogate data and deepen their understanding of the Earth’s climate. Concurrently, we integrated Python into our graduate-level Numerical Weather Prediction course, wherein students use Python resources to visualize and analyze outputs from toy models and the full-physics WRF-ARW model. We began using our The Littlest JupyterHub instance to deploy Jupyter Notebooks and provide a web-based Python programming and visualization environment in support of these courses beginning in Fall 2021 (Numerical Weather Prediction) and continuing to the present (Spring 2023; Introduction to Climate Science and Synoptic Meteorology II). Students and instructors both appreciate having a consistent web-based environment they can access from anywhere, using nearly any device, to complete course assignments and activities, and our Unidata-funded resources will continue to be a part of these courses for years to come. Altogether, over twenty students have been introduced to Unidata and related technologies through these courses.
If you'd like to see how we're using Jupyter resources, the Notebooks developed in support of the Synoptic Meteorology I lab are available in the SynopticMet Github repository. The Jupyter Notebooks developed in support of the Synoptic Meteorology II lab will be available from the same repository shortly after the course’s completion in May 2023.
Jupyter Notebooks developed in support of Numerical Weather Prediction are available on GitHub in two repositories, one containing example notebooks for working with WRF-ARW model data (as is a part of three course assignments and the term project) and one containing example notebooks for working with text-based and GRIB-formatted meteorological data (as is a part of two course assignments).
The Introduction to Climate Science course leverages Jupyter Notebooks provided by Prof. Eli Tziperman at Harvard University as part of his 2022 textbook, Global Warming Science: A Quantitative Introduction to Climate Change and its Consequences. The notebooks are available on the Global Warming Science course site.
To assist other universities who may wish to deploy a local JupyterHub using The
Littlest JupyterHub in support of teaching and/or research, we have prepared a short
guide to doing so; it is available as the appendix to our Equipment Award project
report on the Unidata web site:
Upgrading THREDDS and Deploying JupyterHub at the University of Wisconsin-Milwaukee to Support Education and Research
Please feel free to broadly share this guide, and we welcome any additions, clarifications, or questions!
In addition to its use inside our formal curricula, the Unidata-funded server has been used in our Innovative Weather program, which trains students in the rigors of operational meteorology to provide impact-based forecasts and decision support to public- and private-sector clients across the Midwest United States. Data received through the IDD are locally parsed using automated Python scripts into meteorological analyses routinely used by the student forecasters as part of their forecast-preparation process. Likewise, these scripts themselves were developed by student interns using our local The Littlest JupyterHub instance as a pre-deployment sandbox. Altogether, over ten additional students have been introduced to Unidata and related technologies through these courses.
Unidata Community Equipment Award grants make funds available to colleges and universities to purchase equipment or cloud-based computing services that will enhance their participation in the Unidata program. For additional information on program, visit the Equipment Awards page.