GeoLayers Studio
Keep source data, map context and outputs in one project.
Explore →CASE STUDYEarth
Earth
Observation Change Detection
An applied GeoLayers workflow showing how to move from project evidence to a structured earth observation change detection result.
Browser-based GISGeoAI + spatial analyticsSouth African context
THE DECISION
What changed, where and by how much?
Define the phenomenon being monitored before choosing imagery. Development, vegetation loss, water change and disturbance may require different sensors, dates, indices and thresholds.
IMAGERY
Choose comparable source scenes
Search or load imagery with similar seasonal conditions, suitable cloud cover and consistent geographic coverage. Review the band configuration and create visual composites before calculating change.
DERIVATION
Calculate indicators for each date
Generate the appropriate spectral index or raster expression for both periods. Inspect value distributions and statistics separately so a sensor or cloud artefact is not mistaken for real change.
CHANGE
Compare dates and localise the signal
Calculate absolute or percentage difference and classify gain/loss where appropriate. Use project polygons and zonal statistics to summarise the change by site, municipality, catchment or management unit.
OUTPUT
Move from pixels to a decision-ready result
Publish change layers to the project, combine them with contextual vector data, and create a map/dashboard that separates observed change from interpretation. Use field verification or authoritative ancillary data where decisions depend on the result.
PROJECT INPUTS
Evidence required for a credible analysis
The quality of the result depends on the source evidence and the assumptions attached to it.
- Comparable imagery for the before/after dates
- Cloud and quality information
- Correct band assignments for the chosen index
- Project boundaries or reporting zones
- Context layers for interpreting observed change
DECISION OUTPUTS
What the workflow should produce
A useful case-study result should leave the project with evidence that can be inspected, compared and communicated—not only a final map.
- Raster difference or percentage-change layer
- Gain/loss classification where appropriate
- Summary statistics by polygon
- Areas requiring visual or field verification
- Map products that distinguish observation from interpretation
INTERPRETATION & LIMITATIONS
Separate observed spectral change from the reason it happened
Change detection shows that pixel values or derived indicators changed; it does not automatically explain the cause. Seasonal vegetation, rainfall, sensor differences, cloud/shadow, fire, construction and land management can produce similar signals. The final interpretation should therefore combine temporal context, visual inspection and—where decisions depend on the result—field verification or authoritative supporting data.
- The before/after scenes are comparable enough for the intended question.
- Cloud, shadow and seasonal effects are acknowledged.
- The change threshold is documented rather than chosen only for visual impact.
- Zonal summaries and maps clearly distinguish observed change from interpretation.
RELATED CAPABILITIES
Recreate the workflow with connected GeoLayers tools
The case study pattern can be adapted to your own project data rather than copied as a fixed recipe.