CyberGIS-Jupyter for Water Quarterly Release Announcement (2022-Q2)

Dear CJW users,

We are pleased to announce a new release of the CyberGIS-Jupyter for Water (CJW) platform at https://go.illinois.edu/cybergis-jupyter-water. This release includes several new capabilities and features summarized as follows.

1) CJW moved to a new home. Jetstream-1, an NSF-funded high-performance cloud computing resource where CJW was hosted for the past 3 years, was permanently shut down on July 31, 2022. Its successor, Jetstream-2, which offers much more powerful capabilities, has become the new home of CJW. All existing CJW user data and notebooks have been migrated to Jetstream-2. We do not expect users to experience any change in usage due to this transition but to enjoy a faster and smoother Jupyter environment backed by the latest hardware and cloud technology. In exceptional cases, the previous CJW instance on Jetstream-1 could be accessible upon user request.

2) Improved user experience in CyberGIS-Compute job submission: Have you ever had a long-running job submitted to high-performance computing (HPC) resources but found your Jupyter session died after the browser was idle for too long? The latest CyberGIS-Compute SDK now allows you to reinstate job submission sessions for all previous jobs you submitted. Just switch to the new “Your Jobs” tab page in the user environment and “Restore” the jobs you are interested in. This also gives you a chance to re-download model outputs from previous jobs.

3) WRFHydro model integration supports merging model outputs: A new option “Merge_Output” is added to the WRFHydro workflow supported by CyberGIS-Compute. If enabled, single-timestep NetCDF files can be merged on the “Time” dimension after model execution. Currently supported output types include CHANOBS, LDASOUT, GWOUT, LAKEOUT, RTOUT, and LSMOUT. This optional merging step can reduce data transfer size and speed up post-processing work on CJW. The merged files are put into a separate folder called “Outputs_Merged” alongside the original model outputs. Users can choose to download either or both. Please refer to the example notebook for more information.

4) Enhanced support for user customization to CJW kernels: While CJW has pre-installed a large collection of common libraries and tools to support a suite of hydrologic analysis and modeling workflows, users may still want to install something specific to certain use cases. CJW now allows users to directly use “!pip install XXX” in notebook cells to customize existing kernels. CJW supports flexible additions or changes on a per-kernel basis, which does not affect other existing kernels. Please refer to this example notebook for more information.

5) Updates on CJW backend (kernel, plugin, and bugfix): A new general-purpose kernel, Python3-2022-06, is added, which incorporates a rich set of new geospatial packages. The ‘StickyLand” JupyterLab plugin is installed that allows users to create customizable dashboards and linear notebooks; A bug specific to Apple Safari browser in the OpenWith operation has been fixed.

Please refer to the following resources for details and examples:

Run WRFHydro model on HPC resources using CyberGIS-Compute V2 (updated 2022-07)
https://www.hydroshare.org/resource/cc28d769943046fdac0f9b0c0fc2afc6/

Customization to CJW Kernels with Pip
https://www.hydroshare.org/resource/d18886d2aedf4a2e8c6302165b8fe10f/

CyberGIS-Compute SDK new features
https://cybergis.github.io/cybergis-compute-python-sdk/release-notes.html

CJW 2022-Q2 Release Notes on HydroShare
https://www.hydroshare.org/resource/34b04302d8b34cc6aab826f79b5e3802/

Please let us know if you have any questions or run into any problems (help@cybergis.org). Any feedback would be greatly appreciated.

Best regards,
CyberGIS-Hydro team