9.6.3 — Informal Settlement Analytics — maturity: live
Service Gap Identification
Using multispectral and SAR satellite imagery to locate informal settlements without piped water, grid electricity, paved roads or sanitation infrastructure, at city-wide scale.
Satellite imagery and spectral analytics let city authorities pinpoint exactly where water, sanitation, electricity, and road access are absent in informal settlements — before the next crisis forces their hand.
Municipal planners in rapidly urbanising countries rarely know where service gaps are worst. Ground surveys are slow, expensive and politically shaped; administrative records stop at the formal city boundary. The result is that clinics, water mains and feeder roads get extended to communities with the loudest political voice, not to those with the greatest need. A city of two million informal residents can have its entire service-gap map refreshed in a week using satellite data that no ward councillor can redraw.
The satellite stack layers three signal types. Optical imagery at 50 cm resolution reveals road-surface texture, roof material and building density — proxies for structural permanence and internal access. Nighttime light composites from low-noise VNIR sensors flag electricity absence at block level; a settlement glowing at 0.05 nW/cm²/sr is almost certainly off-grid. SAR coherence and backscatter detect standing water and muddy tracks that confirm absent drainage. Fused in a sovereign GPU pipeline, these signals produce a service-gap index scored 0–100 for each settlement block, exportable directly into the municipality's GIS.
The operational outcome is a prioritised capital-works list that finance ministries and international development lenders can trust. Ward-level gap scores become defensible evidence for budget allocations, World Bank conditional grants and UN-Habitat reporting obligations. Critically, the same pipeline that produces the annual gap map also verifies whether last year's interventions actually reached the intended beneficiaries — feeding directly into §9.6.4 In-Situ Upgrading Verification. Governments that own the pipeline control the narrative; those that rent it discover the vendor has already licensed the same analysis to a competing development consultancy.
Frequently asked
Which satellite data types work best for identifying service gaps in informal settlements?
Very-high-resolution (VHR) optical imagery at 0.3–0.5 m GSD is the primary tool for mapping structure types, road access, and visible infrastructure absence. SAR (Sentinel-1, ICEYE, Capella) adds cloud-penetrating change detection. Multispectral indices (NDVI, NDWI) from Sentinel-2 or Landsat can identify surface-water proximity. A sovereign programme typically fuses all three rather than depending on a single commercial source.
How does satellite-derived service-gap mapping compare in cost and accuracy to household surveys?
World Bank analyses put satellite-derived baselines at 60–75% lower cost than equivalent ground surveys for equivalent spatial coverage. Accuracy for structural proxies runs around 78% without validation; with a stratified 5–10% ground-truth sample accuracy rises to 88–93%, which is sufficient for targeting capital investment priorities. Surveys remain necessary for demand-side data (household income, actual usage behaviour) that imagery cannot see.
Can a national mapping agency run this capability without a commercial imagery subscription?
Yes, with caveats. ESA's Copernicus programme provides free Sentinel-1 and Sentinel-2 data globally, and USGS provides free Landsat archive access. These cover change detection and coarse service-zone modelling at 10–30 m resolution. For the structure-level accuracy (0.5 m) needed to count individual dwellings or map alley widths, a nation either procures its own VHR microsatellite or enters a commercial licensing arrangement — which is exactly the dependency sovereign ownership eliminates.
What machine-learning methods are standard for classifying service-deprived zones from imagery?
Random forest and gradient-boosted classifiers trained on labelled roof-material and road-access features remain the operational workhorses because they handle mixed spectral datasets well and produce interpretable feature-importance outputs. Convolutional neural networks (CNNs) and semantic segmentation models (U-Net variants) deliver higher accuracy on dense urban fabric but require large labelled training datasets and GPU infrastructure. Several national geospatial agencies now maintain hybrid pipelines combining both.
How frequently should a city re-run its service-gap analysis?
Fast-growing cities in Sub-Saharan Africa or South Asia should re-acquire at minimum annually, ideally bi-annually, given documented expansion rates of informal settlements. A sovereign LEO microsatellite constellation with 1–2 day revisit makes continuous monitoring feasible; relying on commercial tasking typically introduces 3–6 month lags due to licensing and procurement cycles.
How does this capability interact with SDG reporting obligations?
SDG Indicator 11.1.1 (proportion of urban population living in slums) requires member states to report to UN-Habitat. Satellite-derived mapping is explicitly recognised by UN-Habitat as a Tier II methodology for this indicator. Nations with their own imagery pipeline can produce annual, sub-national breakdowns rather than the 5-year census-cycle estimates currently submitted by most countries, dramatically improving the policy utility of the data.
What are the data-sovereignty risks of using foreign commercial platforms to host this analysis?
When raw imagery and derived gap-maps are processed and stored on foreign commercial cloud platforms, the host-nation government may face contractual restrictions on redistribution, algorithmic opacity in vendor-supplied AI models, and the risk that a geopolitical event could suspend service. A sovereign programme processes imagery domestically under national data-governance law, shares outputs with municipal authorities on national infrastructure, and retains full audit rights over the methodology.
Which international bodies provide funding or technical assistance to help developing nations build this capability?
The World Bank's GFDRR and Urban Resilience programmes co-finance geospatial infrastructure in lower-income cities. The UN-SPIDER network (via UNOOSA) provides technical advisory on satellite-based gap mapping. ESA's Earth Observation for Sustainable Development (EO4SD) Urban programme has delivered informal-settlement analytics to fourteen African and Asian city governments. FAO's Hand-in-Hand geospatial platform shares adjacent rural-service gap methodology transferable to peri-urban contexts.