13.6.5 — Poverty Analytics — maturity: live

Service Access Inequality Mapping

Quantifying geographic disparities in access to health, education, water and financial services by fusing satellite-derived settlement and infrastructure data with ground truth.

Auto-drafted reference page — content reviewed for shape, individual references not yet link-checked.

Governments routinely claim universal service coverage while communities hours from the nearest clinic or school remain invisible to planners. Administrative data is self-reported by the agencies being evaluated, creating a structural blind spot: the areas least served are precisely those least likely to generate reliable records. Satellite imagery cuts through this by detecting built infrastructure—rooftops, road surfaces, power lines, water points—at sub-5m resolution across entire national territories, including the last-mile settlements that field surveys miss or mislocate.

The satellite stack combines optical and SAR imagery to map settlement extents and infrastructure footprints, while nighttime light composites proxy electrification and economic activity. These layers feed a travel-time model that calculates realistic walk-and-road distances from every populated pixel to the nearest service node—school, health post, ATM, piped water standpipe. The output is a continuous inequality surface, not a district average, exposing the intra-district variance that aggregate indices hide and identifying specific communities breaching acceptable access thresholds.

Operationally, the inequality map becomes the master targeting layer for capital allocation: which clinics to build, which roads to upgrade, which mobile-money agent networks to subsidise. Updated annually, it lets finance ministries and line departments track whether investments are closing or widening the gap. Nations that rent this capability from commercial or donor-funded platforms surrender control over the methodology, the update cadence, and—critically—the politically sensitive finding that some regions have been systematically underserved for decades.

What matters

Travel-time to services, not straight-line distance, is the operationally correct inequality metric—road quality and terrain can triple effective distance.
Sub-district spatial resolution is essential: district-level averages routinely conceal pockets of extreme deprivation that skew the political allocation calculus.
Annual update cadence is the minimum needed to attribute change to specific infrastructure investments rather than demographic drift.
Methodology sovereignty matters: donor-run platforms have withheld or softened inequality findings under diplomatic pressure from recipient-country governments.

Sovereignty score: 8/10 — A nation that does not own its service-access map cannot credibly audit its own development commitments or defend its findings against external political interference.

Donor-controlled platforms have precedent for adjusting inequality surface methodology under bilateral pressure, directly affecting which communities qualify for internationally funded programmes.
Service-access data intersects with internal political settlements—revealing that a particular ethnicity or region has been systematically excluded for decades is intelligence with domestic security implications that no government should hand to a foreign operator.
Commercial vendors offering this as a service can withdraw, reprice or impose export-control restrictions on the derived products, leaving national planning cycles without a consistent baseline at the worst possible moment.
Sovereign ownership enables the government to integrate classified infrastructure data—military installations, planned road corridors, protected area boundaries—that cannot legally or safely be shared with a third-party analytics provider.

Reference architecture

Payload: Multispectral optical imager, 3m GSD, 4-band (RGB + NIR), 20km swath; secondary SAR payload, X-band, 5m stripmap resolution for cloud-persistent settlement detection
Bus class: ESPA-class microsat, 120kg wet, 400W payload power; dual-payload bus hosting optical and SAR instruments on a single platform to minimise constellation size
Orbit: Sun-synchronous LEO at 500–550km altitude; 12-satellite walker constellation delivering 3–4 day full-national-territory revisit; dawn–dusk plane for consistent solar illumination of optical payload
Ground segment: 2-station national network (X-band downlink, S-band TT&C) co-located with national mapping agency and meteorological service; SatNOGS nodes as contingency on UHF/2.4GHz for housekeeping telemetry
Software & data
Latency
Cost band
Lead time

References

Measuring Access to Healthcare Using Satellite Data — World Bank Development Research Group — The World Bank documents how satellite-derived travel-time models outperform administrative records in identifying populations beyond two-hour reach of primary health facilities across Sub-Saharan Africa.
Global Accessibility Map (Oxford Malaria Atlas Project) — Malaria Atlas Project / University of Oxford — The Malaria Atlas Project produced a 1km-resolution global travel-time surface to cities using road networks, land cover and terrain, demonstrating the technique's scalability for national inequality assessment.
Nighttime Light Data for Development — NASA Earth Observatory — NASA describes how VIIRS nighttime light composites serve as a validated proxy for electrification and economic activity, underpinning inequality metrics in data-scarce environments.
High-Resolution Settlement Layer (HRSL) — Meta / CIESIN at Columbia University — CIESIN's High-Resolution Settlement Layer uses commercial satellite imagery to map population distribution at 30m resolution, providing the settlement baseline against which service access gaps are calculated.
SDG Indicator 1.4.1 — Access to Basic Services — UN Statistics Division — The UN SDG framework mandates national reporting on the proportion of the population with access to basic services, explicitly recognising satellite and geospatial methods as valid data sources where ground surveys are insufficient.