Tutorials

These tutorials use a benchmark dataset which provides ready-to-use raster layers and sample points for 10 different regions in Europe, according to tiling system from GeoHarmonizer project.

• 1. Raster Gap Filling
  • 1.1. Temporal Moving Window Median (TMWM)
  • 1.2. Temporal Linear Interpolation (TLI)
  • 1.3. InPainting
  • 1.4. Gapfilled Raster Data
  • 1.5. Time first, space later
• 2. Point Overlay
  • 2.1. Dataset
  • 2.2. Space Overlay
  • 2.3. Space-Time Overlay
  • 2.4. Save to CSV and GeoPackage files
  • 2.5. Overlay Benchmarks
• 3. Land Cover Mapping
  • 3.1. Dataset
  • 3.2. Training
  • 3.3. Predictions
• 4. Land Cover Mapping (Advanced)
  • 4.1. Probability output
  • 4.2. EML (Emsemble Machine Learning) and uncertainties
  • 4.3. Feature Selection
  • 4.4. VotingClassifier
  • 4.5. Saving/Loading the model(s)
• 5. Cloud Optimized GeoTIFF and Tiling Processing
  • 5.1. COG properties
  • 5.2. Tiling processing
  • 5.3. Point query
    • 5.3.1. Parallel query
• 6. Parallel Raster Block Processing
  • 6.1. Aggregation of values within a bounding geometry
  • 6.2. Producing processed raster output within a bounding geometry
• 7. Access to Geo-harmonizer datasets
  • 7.1. Accessing OGC OWS
• 8. Quality control of Geo-harmonizer datasets
  • 8.1. Accessibility test
  • 8.2. Completeness test
  • 8.3. Consistency test
• 9. Sentinel 2 Mosaics
  • 9.1. 1. S2 tile composites
    • 9.1.1. 2. Relative orbits
    • 9.1.2. 3. Final stitching
• 10. SpatioTemporal Asset Catalog
  • 10.1. Populating the catalog
  • 10.2. Accessing the catalog
    • 10.2.1. Data access and download