Comparing predictive analytics, multispectral imagery, ground-movement radar, and L-band radar, and why the evidence behind the map is important for water utilities
Two satellite leak detection providers can deliver maps that look almost identical: a water distribution network overlaid with priority areas for field investigation. Both may call their service satellite-based leak detection, but one map draws on vegetation or ground movement, and another on moisture-related radar measurements.
Those outputs are not interchangeable. To understand what a satellite leak detection solution can tell them, utilities should look beyond the map and into the evidence that created those results.
Four Approaches, Four Types of Outputs
Satellite-based leak detection is not a single technology. Different approaches use varied inputs to narrow the search for hidden water loss, and each answers different questions. Understanding that insight is critical in helping utilities choose the right tool for their water loss program.
| Approach | What it evaluates | What the output represents |
| Predictive analytics | Leak history, pipe attributes, geographic information system (GIS) data, and operational data; sometimes satellite-derived inputs | Where leakage is statistically more likely |
| Multispectral imagery | Surface reflectance and vegetation response | Where surface conditions suggest unusual moisture |
| Interferometric synthetic aperture radar (InSAR) | Changes in ground position over time | Where the ground surface has measurably shifted |
| L-band polarimetric synthetic aperture radar (PolSAR) | Moisture-related radar scattering from vegetation, soil, and, in suitable conditions, the shallow subsurface | Where moisture conditions near buried pipes warrant investigation |
Some solutions combine several of these inputs, which can add useful context. But using satellite-derived data is not the same as observing a present-day, leak-related condition below the surface, and a map of priority locations is only as meaningful as the evidence behind it.
Imaging Methods Detect Different Leak Stages
When water escapes from a buried pipe, it changes moisture conditions in the surrounding soil first, often long before any sign reaches the surface. Depending on ground conditions, the leak may later affect vegetation, surface reflectance, pavement, or ground stability. Some leaks never produce visible surface evidence or measurable ground movement at all.
This shapes how each satellite method performs. Multispectral imaging identifies changes in vegetation and surface conditions, and InSAR detects small movements of the ground surface. Both contribute useful clues. However, rainfall, irrigation, seasonal vegetation change, construction, settlement, and other factors can produce the same observations, so a surface clue alone does not reveal its cause.
Frequency Determines How Deep a Satellite Can Penetrate
Synthetic aperture radar (SAR) transmits microwave signals and analyzes the energy that returns. Each radar band uses a different wavelength, which determines how the signal interacts with vegetation, the soil surface, and the ground below.
X-band generally emphasizes canopy and surface features. C-band is sensitive to vegetation, surface, and near-surface conditions. L-band’s longer wavelength interacts differently with vegetation and, under suitable conditions, reaches moisture in shallow soil. For buried infrastructure, that subsurface interaction is critical, because a leak does not need to surface to waste water.
No satellite “sees” a broken pipe. Soil type, moisture, vegetation, surface cover, depth, and acquisition conditions all influence the measurement. Under the right conditions, L-band PolSAR technology detects leaks up to 10 feet underground across an entire distribution network. The practical question for any vendor is whether its radar data contain information relevant to the conditions it claims to analyze.
Polarization Provides A More Complete Measurement
Frequency and imaging methods are only part of the puzzle. Polarization describes the orientation of the radar energy a satellite transmits and receives. Dual-polarization imagery captures two measurement channels. Full quad-polarization captures four, one for each horizontal (H) and vertical (V) combination, commonly written as HH, HV, VH, and VV. It results in a richer picture of how the signal scatters.
ASTERRA has applied L-band SAR to leak detection since 2013. Today, ASTERRA’s Recover solution pairs full quad-polarimetric L-band data with proprietary processing to analyze moisture-related signatures associated with potable water and wastewater leakage. The additional radar channels help its algorithms interpret those signatures and separate them from other environmental and electromagnetic effects.
What Finding Moisture Actually Means
Moisture near a water main does not automatically mean a water-main leak. Rainfall, irrigation, groundwater, and other environmental sources all produce wet soil. Utilities need to know whether a signal is consistent with the type of water they are trying to find.
Our methodology analyzes dielectric-related radar responses alongside spatial patterns and environmental context to identify signatures consistent with treated-water or wastewater leakage. Trained on over 10 years of data and over 200,000 verified leaks across varied global environments, our model interprets radar and contextual evidence to help utilities decide where to investigate first. It is important to note that it does not measure water chemistry from space, and field confirmation from repair teams is still a must.
In the field, using satellite to prioritize areas of suspicious moisture is imperative to switching a utility’s water management efforts from reactive to proactive. Crews being dispatched because a map shows “wet” will not drastically improve their efficiency in addressing leaks. These teams need areas identified in order of priority, backed by evidence tied to the target infrastructure and water source, as to know where to check first, second, third, and so on.
Combining Several Satellite Datasets Improves Detection
Together, multispectral imagery, InSAR, network records, and artificial intelligence (AI) provide valuable context about surface conditions, ground stability, and historical failure risk. However, what the combined result actually represents can affect the how a utility addresses their suspected leaks.
Several indirect indicators do not, on their own, confirm a current leak-related moisture condition below the surface or identify its source. Utilities should ask how each signal is weighted, whether satellite data were acquired for the target area and time period, which outputs are directly observed versus modeled, and which parts of the method have been validated against field findings.
How Satellite Evidence Guides Results
When datasets are applied and weighed correctly, satellite screening proves its value in the field. For example, the City of Garland, Texas, used ASTERRA’s satellite-based leak detection to guide its crews to 186 verified leaks after two satellite scans. Of those, 141 (76%) were non-surfacing leaks that would likely have stayed hidden. Following those findings, the city reduced its average non-revenue water (NRW) from 120 to 50 million gallons per month, a 58% decrease.
Satellite intelligence strengthens repair teams already using acoustic investigation, correlators, monitoring, and their longstanding, nuanced expertise. While satellite leak detection services vary in what they pick up on, this form of leak inspection remains 3 to 4 times more effective than unguided inspection, and the most useful evaluation measures that difference directly through confirmed leaks, length of pipe inspected, crew time, false leads, and, where measured, actual water savings.
Six Questions to Ask a Satellite Leak Detection Vendor
- What is the primary evidence? Is the result a modeled risk score, a surface indicator, a ground-movement measurement, or a moisture-related radar signal?
- Which sensor and image dates were used? Ask for the satellite or sensor, image dates, radar band, and, where relevant, polarization configuration.
- What does each output represent? Clarify whether a highlighted area indicates a current physical condition, a statistical probability, or another type of observation.
- How are unrelated moisture sources handled? Ask whether the method can distinguish signatures consistent with target-water leakage from rainfall, irrigation, or groundwater.
- Can the methodology be substantiated? Ask for verifiable evidence of the imagery and inputs acquired and analyzed, plus a clear description of how the method was validated.
- What field performance has been demonstrated? Compare confirmed outcomes on comparable networks, not only total points of interest (POIs) or headline claims.
Frequently Asked Questions
Can Satellites Detect Underground Water Leaks?
Yes. Satellites can identify conditions associated with hidden leakage and show crews where to investigate first. L-band PolSAR technology, for example, detects leaks up to 10 feet underground across an entire distribution network. Field crews then confirm and pinpoint each leak before any excavation.
What Is the Difference Between InSAR and L-Band SAR?
InSAR is a radar-processing technique that measures ground displacement over time. L-band refers to a radar frequency, and L-band polarimetric SAR can be analyzed for moisture-related scattering. Because the two terms describe different properties, they are not mutually exclusive.
Can Satellite Detection Tell Potable-Water Leakage From Groundwater?
ASTERRA’s methodology identifies radar signatures consistent with a target water type by analyzing dielectric-related responses alongside environmental context. It is an algorithmic inference that field crews validate, not a direct satellite test of water composition.
