Skip to content

Afternoon — GeoLibre & Exercise 3

Day 3 · 13:30–16:00 · Module 4

Session 8 — Spatial Analysis using GeoLibre (13:30–14:30)

What is GeoLibre?

GeoLibre is a web-based GIS viewer and analysis tool — no installation needed.

GeoLibre can:

  • Open spatial files (KML, GeoJSON, GeoPackage, Shapefile ZIP)
  • Draw points, lines, and polygons on a map
  • Create buffers around features
  • Overlay multiple data layers
  • Display population, land cover, and other raster layers
  • Export maps as PNG images (via Print)
  • Compute basic statistics within a drawn area

Worked example — downloading and using a boundary from GADM

The session walked through a complete workflow using Uganda as an example:

  1. Go to gadm.org → click Data → select Uganda → download as GeoJSON
  2. The downloaded file contains all districts in Uganda as separate polygons
  3. Open GeoLibre → load the GeoJSON file
  4. Select one district from the list → isolate it → export as a separate GeoJSON
  5. Import the district boundary into eToolkit → generate a report for that district

Due diligence on GADM boundaries

Always verify: how many districts does the official government statistics say Uganda has? If GADM says 136 rows and the government says 135 districts, investigate before using in a formal document. Don't accept external boundary data uncritically.

Admin boundary sources

gadm.org — Global Administrative Areas

  • ~500,000 administrative units worldwide
  • Version 4.1 (Version 5 planned for early 2026)
  • Country → ADM0 (national), ADM1 (province/state), ADM2 (district/county)
  • Formats: GeoJSON, Shapefile, KMZ, GeoPackage
  • Free download, no registration

data.humdata.org — Humanitarian Data Exchange (UN OCHA)

  • Large collection of country-level spatial and statistical data: poverty, population, census data, health facilities, education facilities
  • Not fully structured — quality varies by country and dataset
  • Search by country → filter by file type (Shapefile, GeoJSON, CSV)
  • Used in training: download population raster → open in GeoLibre → compute population in a project buffer

geoboundaries.org/simplifiedDownloads.html

  • Open admin boundaries maintained by the community
  • Simplified downloads page: quick access by country and level
  • GeoJSON format — ideal for web tools

GeoLibre limitations

When to use QGIS instead

GeoLibre is web-based — large raster datasets (e.g. a national population grid) can be slow to process. For: - Heavy raster statistics → use QGIS (free, desktop) - Complex spatial queries on large datasets → use QGIS or Google Earth Engine - Quick vector analysis (buffers, overlays, area calculations) on district-scale data → GeoLibre is fast enough

Exercise 3 — Monitoring Indicators for Project Assessment (14:30–16:00)

Objective: Apply EO and spatial analysis to a monitoring or evaluation scenario for a project area you work on or know well.

Tools: eToolkit + GeoLibre + GADM/HDX (as needed)

Part A — Create a before/after comparison in eToolkit

  1. Return to your Exercise 1 or 2 area in the eToolkit.
  2. Run the analysis with the current date range (your "after" snapshot).
  3. Re-run with a time window 3–5 years earlier (your "before" baseline).
  4. Compare the two reports:
    • LULC: has land use changed?
    • Vegetation: is green cover or biomass higher or lower? (ETa or LULC change)
    • Water: has surface water extent changed?
  5. Write one sentence describing the most significant change you observe.

Part B — Spatial analysis in GeoLibre

  1. Download a district or admin boundary for your project area from GADM or GeoBoundaries.
  2. Open it in GeoLibre.
  3. Draw a 1–5 km buffer around a key feature (a road, a school, a dam, an irrigation intake point).
  4. Explore what is inside the buffer: land use, settlement density, visible population.
  5. Export the map as a PNG (Print → Download).

Expected result

  • A before/after description (one paragraph) showing change in your project area
  • A GeoLibre map image showing the project boundary and buffer zone
  • A completed row in the Use-Case Selection Form for the indicator you found most useful

Troubleshooting

  • GADM downloads as ZIP → unzip it first; then load the .geojson or .shp file in GeoLibre
  • GeoLibre is slow with large national boundaries → select only the relevant district and export it before doing analysis
  • eToolkit time-window comparison → check that the start/end dates in both runs are clearly different; the platform processes historical data going back to 2000

Continue to Day 4 — Case Studies & Closing