Second [Virtual] ICESat-2 Cryospheric Hackweek Facilitated by the University of Washington

The 2020 ICESat-2 Cryospheric Science Hackweek was the first virtual Hackweek held by the University of Washington. While originally planned as a five-day, in-person workshop, the event was shifted to a fully virtual/remote setting in light of stay-at-home orders and travel restrictions in place to curb the spread of COVID-19.

To accomodate multiple time zones and limit the daily duration of online tutorial sessions, the event was spread out over the course of ten days. The first week had three half-days of interactive tutorials/lectures. The second week had four days that included some interactive tutorials/lectures and scheduled times where instructors were available to help participants with a facilitated exploration of datasets and hands-on software development.

This June 2020 event resulted in the production of a series of tutorials, developed by volunteer instructors and presented during the event. During the actual Hackweek, teams of researchers and data scientists developed a series of interesting projects related to their interests/research.

Tutorials Zenodo badge

The tutorials from this event live within a dedicated GitHub repository and are published on Zenodo. You can run the tutorials by following the instructions here. The published tutorial repo also includes links to presentation slides and videos of the recorded presentations.

Tutorial Topics:

1. Introductory Session 1. ICESat-2 Mission: Satellite, Sensor, and Data 1. Git and GitHub 1. Jupyter and iPython 1. Geospatial Analysis with Python 1. Introduction to ICESat-2 Sea Ice and Land Ice Products and Data Access 1. Programmatic ICESat-2 data access 1. Introduction to HDF5 and ICESat-2 data files 1. Land ice applications 1. Sea ice applications 1. Science data generation 1. Machine learning

Projects

Though in many cases preliminary, these project repositories can provide useful starting points to develop effective cryospheric workflows where more formal examples and functionality have not yet been developed.

  • icepyx

    • Contributions to icepyx included new example notebooks, packaging on Python Package Index, installation instructions, and automated code coverage reports.

Sea Ice

  • leading to phytoplankton

    • Obtain and visualize coincident ICESat-2, Sentinal-2, and Argo Float data

    • Many members of this group still meet regularly (2 years out!) and are creating a template to add new coincident datasets to icepyx.

    • Group members (including newer members) contribute to icepyx as collaborative developers and code reviewers.

  • overcast

    • Build tools to merge data and explore the effects of sea ice leads on clouds in the Arctic

Glaciers and Ice Sheets

  • Seasonal Snow

    • Compare ICESat-2 data with high resolution DEMs over complex (mountainous) terrain

  • unsupervised

    • unsupervised surface classification of ATL03 photons

  • FirnAndMelt

  • CloudMask

    • Fetch, classify, and label ICESat-2 data

    • Still an ongoing collaboration??

  • crossovers

    • processing of non-overlapping ICESat-2 tracks

  • surface_velocity

    • Infer surface ice velocity from repeat passes and across beams.

    • Continued work resulted in a poster at AGU Fall Meeting 2020

  • Assimilation

    • Compare ICESat-2 elevations with multiple DEM raster data types.

    • Quantify errors and compare results regionally

    • Contributed additional authentication methods to icepyx and provided initial code for what eventually became the Visualization module.