Skip to content
Friedrich Knuth edited this page Jun 21, 2019 · 29 revisions

Topohack 2019


Taking ICESat-2 to the mountains: a workflow using satellite laser altimetry to resolve topography over complex terrain

Background

Science Objectives

  • Compare elevation profiles from ICESat-2 ATL06 products over complex terrain with existing high-resolution DEM datasets
  • Compare Snow-off DEM over Grand Mesa with winter (October to February) IceSat-2 collects.

Data Science Objectives

  • Learn how to collaborate as a team via GitHub
  • Explore Python data science ecosystem
  • Learn how to work on JupyterHub and pull in remote data
  • Create simplified workflow to extract ICESat-2 ATL06 data
  • Build a pip installable library with convenient data access and wrangling functions
  • Explore Python data visualization and analysis tools

GitHub Data Management

Folders

  • topolib: main library with custom python functions built by the team
  • scripts: utilities to be run from command line
  • notebooks: example notebooks to use as a reference
  • images: static figures and images to be pulled into notebooks
  • contributors: scratch space for team members to work without risk of merge conflict

Files

  • .gitignore: file that tells git to ignore certain file types
  • LICENSE: default licence
  • README.md: landing page
  • setup.py: utility to install topolib library

Other

  • To avoid commit large amounts of data to the github repo, common file types (.h5, .tif, .laz) are added to the gitignore.
  • When possible, data files should be saved or moved outside the repository. This further reduces the risk of committing data to GitHub and also makes the data more easily accessible across the system.

Daily Agenda and Objectives

Monday

  • Explore OpenAltimetry and utility with multiple locations and time periods (All)

Tuesday

  • Introductions (All)
  • Practice collaborating with git (All)
  • Add and present utility to download validation data staged with Google Drive storage (Friedrich)
  • Create distilled example of data-access tutorial (Michelle)
  • Discuss general objectives (All)

Wednesday

  • Checkin and touch base on objectives for the day/week (All)
  • Organize individual folders into a single folder (All)
  • Merge all open pull requests (All)
  • Make sure everyone is up to date with the latest changes (All)
  • Work with data-access notebooks to download ICESat-2 data using spatial and temporal filters (All)
  • Explore and understand ATL06 variables, flags, and theoretical basis (All)
  • Learn how to subset ATL06 data based on flags using tutorial notebooks (All)
  • Start creating a python library to make it easier to download by spatial and temporal filters (Joe)
  • Work with HDF5 files as xarray datasets and/or geopandas/shapely polygons ( )
  • Plot profile of ATL06 latitude, longitude, h_li to elevations extracted from reference DEMs (Friedrich)
  • Get classified map of Grand Mesa from WorldView imagery, download ATL03 and ATL08 data for comparison (Michelle)

Thursday

  • Data Filtering (Justin/Mariama)
  • Improve profiling code, chip in any geospatial problem in general (Shashank, Friedrich)
  • Clean functions, make everything usable by all (Joe, Justin, Friedrich)
  • Evaluate/Compare the topography resolved by ICESat-2 profiles along steep mountains with topographic profiles returned from high-resolution DEMs
  • Get a sense of snow accumulation (depth) by comparing Snow-off DEM over Grand Mesa with winter (October to February) IceSat-2 collects.

Friday

  • Summarize progress and next steps

Additional Objectives

  • Intersect ICESat-2 data with RGI glacier polygons to get a sense of bare ground coverage near glaciers.
  • Generate DEM for GM using ASP and compare with ATL06

Study Sites and Objectives (as converged on Wednesday)

  • Grand Mesa (Michelle/HP)
  • ARTIC DEM (Will)
  • HMA (All points within glaciers, plot the distribution) (Vibhor)
  • Mt. Baker (Friedrich)
  • US (Comparison with NED) (Amy)

Methods/Planned Approach

Tools/Libraries

numpy
scipy
rasterio
pandas
geopandas

Collaborators

Vibhor Agarwal
Michelle Hu
Friedrich Knuth
HP Marshall
Justin Pflug
Mariama Dryak
Will Kochtitzky

Team Lead:

Shashank Bhushan

Data Science Leads:

Joe Meyer
Amy Steicker


References

ATL06 technical references