Landsat and Sentinel-2 datalibs

[thabbott]

Changes

Features

• Add Landsat and Sentinel datalibs for searching, retrieving, and visualizing Landsat 8-9 and Sentinel-2 imagery. The datalibs include:
  • Tools for querying Landsat and Sentinel-2 imagery for intersections with user-defined regions (landsat.query, sentinel.query) or flights (landsat.intersect, sentinel.intersect). These tools use BigQuery tables and require a Google Cloud Platform account with access to the BigQuery API.
  • Tools for downloading and visualizing imagery from Landsat (Landsat) and Sentinel-2 (Sentinel). These tools retrieve data anonymously from Google Cloud Platform storage buckets and can be used without a Google Cloud Platform account.
• Add tutorial notebooks demonstrating how to use Landsat and Sentinel datalibs to find flights in high-resolution satellite imagery.

Internals

• Add unit tests for Landsat and Sentinel search tools and datalibs, but disable any tests that retrieve imagery data when running tests in GitHub Actions to limit resource consumption in GitHub runners. Users can disable these unit tests locally by setting the environment variable GITHUB_ACTIONS=true.

Tests

• [x] QC test passes locally (make test)
• [x] CI tests pass

Reviewer

@mlshapiro

[review-notebook-app[bot]]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[thabbott]

I think your re-organization makes sense! Moving the generic search and query functions to _leo_utils should help make it clear to users that they (in almost all cases) want to use e.g. landsat.search and not _leo_utils.search.

I added some quotes about get calls taking a while to notebooks in https://github.com/contrailcirrus/pycontrails/pull/200/commits/8dcd483bf3c3254c28aba6a21cc3b44d0ce6d22b. (They've been pretty quick for me, but that's been running on a VM with fast transfer speeds from google cloud buckets. I suspect you were running locally?)