New Mexico is turning to satellites and artificial intelligence to tackle a costly problem of drinking water leaking from underground pipes.
New Mexico’s Environment Department has launched LeakTracer, a statewide program using satellite imagery, artificial intelligence, and field inspections to detect underground drinking water leaks, particularly benefiting smaller and rural utilities with limited monitoring resources. The program analyzes subsurface moisture using synthetic aperture radar, distinguishing treated drinking water from other sources, then directs field inspectors to verify suspected leak locations identified through comparison with utility pipeline maps.
After its first year, the program worked with 28 utilities, inspecting 181 miles of pipeline and identifying 216 leaks. If repaired, these could recover an estimated 898,000 gallons daily and reduce production costs by about $505,000 annually, though these figures remain projections pending actual repairs. While the state covers detection costs, utilities must fund repairs themselves, posing challenges for resource-limited systems despite available infrastructure grants and loans.
Through LeakTracer, a statewide initiative led by the New Mexico Environment Department (NMED), the state is helping water utilities find leaks beneath streets, homes and other infrastructure before they become visible at the surface.
The goal is straightforward: identify where treated drinking water is escaping, help utilities prioritize repairs and prevent valuable water from being wasted.
The need for such technology is particularly important for smaller towns and rural communities, where water utilities may have limited resources to monitor extensive pipe networks and maintain aging infrastructure.
LeakTracer combines satellite imagery, artificial intelligence and on-the-ground inspections to help utilities detect leaks that might otherwise remain undiscovered. The initiative also supports New Mexico’s 50-Year Water Action Plan, which calls for conserving water and reducing losses from public water systems.
Why use satellites and artificial intelligence?
Traditional leak detection can be challenging when water escapes from pipes buried underground. A leak may persist without producing obvious surface signs, allowing treated water to disappear before utility crews can identify the source.
LeakTracer offers another way to locate these problems by analyzing underground moisture patterns from space.
The state works with ASTERRA, a San Diego-based company specializing in satellite-based leak detection. Its technology uses L-band synthetic aperture radar (SAR), which analyzes signals reflected from the Earth’s surface and subsurface.
According to NMED Deputy Director of Communications Breanna Anderson, the radar can penetrate up to 10 feet underground.
The technology looks for patterns associated with moisture from treated drinking water. Artificial intelligence then helps distinguish those patterns from moisture caused by rainwater, groundwater and other sources.
Water treated with chlorine and similar sanitizing agents can have a different signature from naturally occurring water, Anderson said.
“What makes the LeakTracer program stand out nationally is that this allows the identification of potable water separate from rainwater, groundwater, or other surrounding water sources,” she said.
That distinction is important because detecting moisture alone does not necessarily mean a water pipe is leaking.
After the satellite data is analyzed, the locations flagged as potential leaks are compared with geographic information system (GIS) maps of utility pipe networks. Field inspectors then investigate the suspected locations, helping utilities determine where repairs may be needed.
The technology does not replace inspections or repair crews. Instead, it helps direct their attention to places where leaks are more likely to occur.
Millions of gallons could be saved
The state’s decision to expand LeakTracer statewide followed a pilot program in five communities. According to Anderson, the results provided the confidence needed to move beyond small-scale testing.
During its first year, the broader initiative worked with 28 water utilities, inspected 181 miles of distribution pipeline and identified 216 leaks.
The findings included 45 leaks on main lines, 54 on service lines and 117 on customer-owned lines.
If all 216 leaks are repaired, NMED estimates the work could recover approximately 898,000 gallons of drinking water per day.
That amounts to roughly 328 million gallons annually — equivalent to the yearly water use of about 6,400 New Mexican households.
The potential financial benefit is also significant. Repairing all the identified leaks could reduce water production costs by approximately $505,000 annually, according to the department’s Year One Summary.
These savings could give utilities additional resources for infrastructure upgrades, further repairs and other operating needs.
But the figures are projections, not confirmed results. The estimated water recovery and financial savings depend on all identified leaks being repaired and producing the expected reductions in water loss.
NMED is developing a process to follow up with utilities, document completed repairs and calculate actual water savings.
The challenge: paying for repairs
Although New Mexico covers the cost of satellite-guided detection and on-site verification through LeakTracer, participating utilities remain responsible for paying to fix the leaks.
That can be a significant hurdle for smaller and rural systems, where budgets may already be stretched by maintenance and infrastructure needs.
Anderson identified the cost and logistics of excavation and repair as major challenges for participating communities.
The department’s Rural Infrastructure Program can help eligible water systems seek grants and loans for infrastructure projects.
The state launched the broader statewide effort in January 2026, offering the service to qualifying public water systems serving communities with fewer than 20,000 people.
As of July 2026, 43 utilities had signed up. By July 1, the start of fiscal year 2027, more than 60 had signed up or received services, Anderson said.
NMED expects another 50 to 60 utilities to participate during the program’s second year.
The department is also exploring whether the service could eventually reach larger public utilities, private systems and tribal communities.
Can AI make leak detection better?
The technology’s effectiveness could improve as it analyzes more data.
Anderson said the AI models can be trained using additional information gathered through the program, potentially helping them become better at identifying patterns associated with underground drinking water leaks.
“AI models are trained on data. The more data you have, the better the algorithm is,” she said. “So, as data is collected, it is used to train the model to make it better.”
The approach could help utilities identify suspected leaks more efficiently, although field verification and follow-up repairs remain essential to realizing the benefits.
For New Mexico, the larger objective is to reduce avoidable water losses as the state pursues long-term conservation goals.
LeakTracer offers utilities a new way to find water that would otherwise continue disappearing underground. Its ultimate success, however, will depend not only on how many leaks satellites and AI can identify, but also on whether communities have the resources to repair them.
The technology can point utilities toward the problem. Turning those discoveries into lasting water savings will require action on the ground.
