11.6.1 — Utility Infrastructure — maturity: live

Water Network Leak Indicators

Using satellite-derived soil moisture, surface deformation and thermal anomalies to flag probable leak zones in buried water distribution networks before catastrophic failure.

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

Water utilities in most countries lose between 20% and 40% of treated water to leakage before it reaches a tap. Ground crews with acoustic loggers can only survey so many kilometres per year, and the backlog of undetected leaks grows faster than inspection capacity. Satellite observation changes the economics: persistent wide-area coverage can rank every pipe corridor by leak probability and direct ground crews only where the evidence is strongest.

Three complementary payloads do the work. Synthetic aperture radar interferometry (InSAR) detects millimetre-scale ground subsidence that accumulates above slow, chronic leaks. Thermal infrared imagery spots the cooler surface signatures of water migrating upward through soil. Multispectral bands catch the anomalous vegetation greenness that thrives above persistent moisture. Combined, these form a probabilistic leak-likelihood layer updated on each satellite pass.

The operational outcome is a ranked work-order queue delivered to field maintenance teams each morning. Utilities that have piloted similar approaches report a 30–50% reduction in physical survey kilometres while actually increasing the detection rate of significant leaks. For a sovereign operator, the network map and loss statistics never leave national infrastructure — a material consideration when water security is a strategic priority.

What matters

InSAR phase coherence degrades over vegetation; bare-soil or paved corridors give the clearest subsidence signal, and urban networks are precisely where leaks are most expensive.
A single undetected trunk-main failure can void months of water treatment investment and trigger a public health emergency within 48 hours.
Commercial InSAR services are export-controlled or subscription-gated, meaning a foreign vendor can suspend access during a dispute or crisis at exactly the moment national infrastructure is most stressed.
Revisit frequency drives detection latency: a 6-day repeat misses fast-developing failures, so a national constellation targeted at domestic pipe corridors can be tasked daily.

Sovereignty score: 8/10 — A nation's water distribution network is critical national infrastructure whose vulnerability map must never reside on a foreign vendor's servers.

Leak-location data is, in effect, a detailed map of pipe condition and pressure weak-points; handing it to a commercial foreign provider creates an intelligence exposure that adversaries could exploit to plan sabotage or coercive pressure during a dispute.
Export-control regimes and vendor contractual clauses routinely allow suspension of InSAR data services, meaning a utility could lose its primary leak-detection feed at the precise moment of a national emergency or political tension.
Water loss rates and infrastructure condition are politically sensitive; a sovereign system lets the government control when, how and whether that data is disclosed to regulators, media or foreign partners rather than having it inferred from a vendor's archive.

Reference architecture

Payload: C-band SAR (5.405 GHz) for InSAR subsidence mapping at 5m resolution, 80km swath; secondary longwave thermal infrared (8–12 µm) channel at 30m resolution for surface moisture anomaly detection; optional 4-band multispectral (B, G, R, NIR) at 10m for vegetation index cross-check
Bus class: ESPA-class microsat, 120–160 kg, 600 W payload power; SAR antenna deployable to 2m × 0.5m planar array; thermal channel integrated as a compact pushbroom sensor on the same bus
Orbit: Sun-synchronous LEO at 520–560 km altitude; 6-satellite walker constellation providing 1–2 day revisit over national territory; dawn-dusk orbit preferred for stable solar power and consistent thermal acquisition geometry
Ground segment: 2-station national X-band downlink network (primary at capital, secondary at geographic antipode for resilience); S-band TT&C; on-site Level-0 ingest servers air-gapped from public internet; SatNOGS UHF beacon for housekeeping backup
Software & data
Latency
Cost band
Lead time

References

Non-Revenue Water — The Global Picture (IWA) — The International Water Association estimates global non-revenue water losses exceed 346 billion litres per day, costing utilities approximately USD 39 billion annually. Reducing physical losses is identified as the single highest-return intervention for utilities worldwide.
Sentinel-1 InSAR Applications for Infrastructure Monitoring (ESA) — ESA documents the use of Sentinel-1 C-band SAR interferometry for detecting ground deformation associated with underground infrastructure stress, including water and sewer networks. Sub-centimetre displacement sensitivity is achievable at 20m ground resolution.
Remote Sensing for Water Utility Asset Management (USGS Water Resources) — USGS research confirms that multi-temporal thermal and multispectral satellite imagery can identify surface soil moisture anomalies correlated with subsurface water loss events. The approach is particularly effective in arid and semi-arid regions where background moisture contrast is high.
Space-Based Earth Observation for Urban Infrastructure (World Bank) — The World Bank's urban infrastructure programme highlights satellite-derived deformation and thermal datasets as cost-effective tools for low- and middle-income country utilities that lack the field inspection workforce to survey large network footprints. Sovereign data access is flagged as a governance prerequisite for national water security planning.
Water Loss Control Manual — Leak Detection Technologies (EPA) — The US EPA's Water Loss Control guidance notes emerging remote sensing methods as a complement to acoustic and pressure-based detection, particularly for prioritising survey zones across large distribution systems. It acknowledges that traditional methods cannot scale economically to full-network coverage.