Remote Sensing Labs

Remote Sensing Labs Roadmap

Remote Sensing Labs is shifting from a product-specific course into a broader learning resource for satellite imagery, geospatial analysis, and reproducible remote sensing workflows.

The existing Google Earth Engine material remains useful and should stay available. The goal is not to delete that track. The goal is to make Earth Engine one implementation option inside a concept-first site.

Current State

  • The site has a complete Earth Engine-oriented tutorial sequence.
  • The strongest content is conceptual: digital imagery, resolution, projections, visualization, indices, classification, and time series.
  • Many examples are tied to Earth Engine APIs and dataset IDs.
  • The source imagery is mostly screenshots and examples captured from the original course.

Direction

The site should organize around durable remote sensing concepts:

  1. Image structure, bands, and raster data
  2. Coordinate reference systems, projections, and scale
  3. Visualization and band combinations
  4. Spectral indices and transformations
  5. Classification, validation, and uncertainty
  6. Time series and change detection
  7. Data access, reproducibility, and workflow design

Each concept can then support one or more implementation paths:

  • Earth Engine for fast browser-based and cloud-scale examples
  • Python with rasterio, rioxarray, xarray, geopandas, shapely, and scikit-learn
  • STAC catalogs and cloud-optimized GeoTIFFs for modern imagery access
  • QGIS or other desktop workflows where visual inspection matters
  • Small licensed very high resolution examples only where public imagery cannot show the concept cleanly

Migration Rules

  • Preserve useful existing URLs.
  • Fix broken links before rewriting lesson code.
  • Modernize one page at a time.
  • Keep the original Earth Engine example until the replacement has been tested.
  • Prefer public/open imagery for baseline lessons.
  • Add paid or licensed imagery only as a documented proof of concept, not as a dependency for the whole course.

First Pages To Rework

Start with the pages where the concepts are broadly useful and the platform dependency is easiest to loosen:

  1. Digital Images
  2. Resolution
  3. Visualizing Images
  4. Indices
  5. Classification

Time series and local environment pages should come later because they require stronger decisions about data storage, compute environment, and notebook tooling.