Link to the code: Zae-Project / arkspace-core

Listening for Falling Satellites: How Seismometers Are Becoming Space Debris Trackers

The global network of seismometers deployed to detect earthquakes has a secondary sensitivity most engineers never designed into it. When spent rocket stages, defunct satellites, or large debris fragments re-enter Earth’s atmosphere at orbital velocity, they produce infrasound and seismic signals that propagate across the planet, detectable by instruments originally installed for geophysics.

A paper published in Science in January 2026, led by researchers from Imperial College London and Johns Hopkins University, documents the first systematic demonstration that this secondary sensitivity can be exploited for operational debris tracking. The implications extend well beyond re-entry monitoring. The technique fills a surveillance gap that existing radar and optical systems cannot close, and it does so using infrastructure already operational at hundreds of stations worldwide.

Why the Tracking Gap Exists

The debris environment in Low Earth Orbit is characterized by objects ranging from intact defunct satellites several meters across down to paint flecks and propellant residue measured in micrometers. Surveillance capability correlates with object size in a way that leaves the most operationally dangerous size range least well characterized.

Objects larger than 10 centimeters in LEO are tracked by the U.S. Space Surveillance Network, which operates ground-based radar and optical systems. The current catalog contains approximately 40,000 objects in this size range. These objects can be assigned orbital elements precise enough to generate conjunction warnings for operational satellites.

Objects below 1 centimeter have limited ability to critically damage a satellite’s structural integrity in most collision scenarios, though they can cause surface erosion and sensor degradation over time.

The 1-10 centimeter range is where the problem sits. These objects are too small for current radar networks to reliably track to catalog precision, but large enough to critically damage or destroy operational satellites. A 1-centimeter aluminum sphere at orbital relative velocities carries approximately the same kinetic energy as a hand grenade. There are estimated hundreds of thousands of objects in this size range, precisely located for only a small fraction.

The space debris cleanup missions that launched in 2026 target large, precisely catalogued objects specifically because those are the ones that can be tracked to rendezvous precision. The removal of small, uncatalogued debris remains technically out of reach. Tracking those objects at all, even without removal capability, would allow better statistical modeling of collision probability and more accurate avoidance maneuvering.

How Re-entry Acoustics Work

When an object re-enters Earth’s atmosphere, it decelerates from orbital velocity (7-8 km/s for LEO objects) to subsonic speed over a period of minutes. The energy released during this deceleration is enormous, comparable in some cases to kilotons of TNT equivalent. A substantial fraction of this energy propagates as infrasound waves, acoustic signals below the 20 Hz threshold of human hearing.

Infrasound propagates efficiently through the lower atmosphere and can travel thousands of kilometers from the source. Seismometers, which detect ground motion caused by pressure waves coupling into the solid Earth from the atmosphere, can register these signals even at continental distances from the re-entry track.

The Imperial College London and Johns Hopkins team analyzed re-entry events from the U.S. Space Surveillance Network catalog where the re-entry trajectory was independently verified, then searched the waveform archives of 247 seismometer stations from the International Federation of Digital Seismograph Networks for corresponding signals.

Their results established that re-entry signatures from objects as small as 100 kilograms are detectable with high confidence at distances up to several thousand kilometers when atmospheric conditions are favorable. For larger re-entries, the triangulation precision from multiple stations could determine impact locations to within tens of kilometers within minutes of the event.

What the Technique Actually Enables

The immediate application documented in the paper is improved re-entry tracking for already-catalogued objects in the final phase of orbital decay, when radar tracking precision degrades because the object’s behavior becomes increasingly dependent on atmospheric density and attitude, which are difficult to model precisely.

The broader application, which the authors describe but do not fully demonstrate, is detection of uncatalogued re-entries. If a debris fragment too small for radar catalog tracking re-enters from a size range of interest, the seismometer network would detect it and could place it in the re-entry record with a rough time and geographic location, even without a corresponding catalog entry. Over thousands of events, this creates an empirical database of the uncatalogued debris population that is currently characterized only through statistical inference and laboratory simulations of hypervelocity fragment distributions.

The World Economic Forum’s January 2026 “Call to Action” on space debris identifies improved characterization of the small debris population as a prerequisite for accurate economic risk modeling. Insurance underwriters and satellite operators cannot price orbital risk accurately without better statistics on the 1-10 cm debris population. The seismometer technique provides a method to accumulate those statistics without new space hardware.

Integration with Autonomous Constellation Management

The operational satellite industry has developed increasingly automated responses to conjunction warnings. SpaceX’s Starlink constellation performed nearly 50,000 avoidance maneuvers in 2023, driven by its autonomous collision avoidance system. The distributed AI coordination protocols being deployed across satellite constellations in 2026 represent the next level of this capability, with multi-satellite negotiation of maneuvering decisions to minimize fuel consumption while maintaining coverage.

These systems are fundamentally limited by the quality of the debris catalog they query. A conjunction warning system that cannot track objects in the 1-10 cm range cannot protect against collisions with those objects. The value of the seismometer technique is not collision avoidance, since re-entry events are not collision events. It is statistical: building a better model of where the debris is, how it is distributed by altitude and inclination, and how quickly the environment is evolving.

Better statistical models enable more appropriate risk-based maneuvering thresholds. The current thresholds reflect uncertainty about the uncatalogued debris population. With better statistics, operators could distinguish between high-density orbital regimes that genuinely warrant frequent maneuvers and lower-risk zones where conservative maneuvering policies waste fuel unnecessarily.

The software-defined satellite architectures that allow real-time reconfiguration of operational behaviors could integrate debris statistical updates directly into constellation maneuvering policy, adjusting risk thresholds dynamically as the empirical picture of each orbital cell improves.

Limits of the Technique

The authors are careful to document where the seismometer approach fails. It is a passive sensor of re-entry events, not an orbital tracking system. It cannot provide the continuous orbit determination needed for conjunction warning. It can tell you that something fell; it cannot tell you what orbit it was in before it did.

Atmospheric coupling efficiency varies substantially with season, weather, and stratospheric wind profiles. The atmospheric waveguide that channels infrasound over long distances is highly variable. Events detected with confidence under one set of atmospheric conditions may generate no detectable signal at the same station network under different conditions.

The technique also requires calibration with events of known source characteristics to validate size and mass estimates from seismic signal amplitude. This calibration database is currently limited to the events where catalog objects re-entered in ways that were well-monitored by other means. Extending confidence to uncatalogued events requires larger calibration sets and more refined source models.

Relevance for Future Debris Tracking Architecture

The reason this research matters for orbital infrastructure planning is that the constellation growth trajectory of the late 2020s and 2030s, driven by programs from SpaceX, Amazon, the European Union, and China, makes improved debris tracking increasingly urgent. More operational satellites means more maneuvering activity, more conjunction events, and more pressure on tracking infrastructure that was designed for a far sparser orbital environment.

The World Economic Forum’s strategic risk framing treats space debris as a cost externality that the industry has collectively underinvested in monitoring and mitigation. The seismometer technique represents a low-cost, high-leverage way to extend monitoring capability using existing infrastructure. It doesn’t require a new space mission or a new ground radar system. It requires data fusion and signal processing applied to a sensor network already deployed for other purposes.

The operational threshold for usefulness is improving the statistical characterization of the uncatalogued debris population to the point where the 1-10 cm size range can be modeled with confidence intervals small enough for actuarial applications, specifically for orbital debris insurance pricing and for setting maneuvering policy thresholds in autonomous constellation management systems.

The on-orbit servicing missions validating proximity operations in 2026 will eventually extend to small debris targeting as the technology matures. Those missions will need improved target characterization to select removal candidates efficiently. A seismometer-derived database of re-entry events provides part of that characterization by tracking the population’s decay rate in the size ranges of removal interest.

Debris that falls is debris that no longer threatens operational satellites. Understanding the natural decay rate of the uncatalogued population, and confirming that models of it match observations, is a necessary step toward understanding when active debris removal is necessary rather than merely beneficial.

Path Forward

The Imperial College London and Johns Hopkins paper establishes proof of concept. The next steps they identify are: expanding the calibration dataset with more re-entry events, improving atmospheric propagation models to extend detection range under variable conditions, and developing a real-time fusion pipeline that ingests seismometer waveforms alongside radar tracking data to produce integrated re-entry estimates within hours of events.

The International Monitoring System operated by the Comprehensive Nuclear-Test-Ban Treaty Organization maintains an infrasound array network specifically designed for atmospheric nuclear test detection. Its 60 operational stations provide global coverage and calibrated sensitivity that complements the seismometer network. Data sharing between these two global sensor networks could substantially improve detection confidence and geographic coverage, particularly over ocean areas where the seismometer network is sparse.

Turning a seismology network into a debris tracking system is not a transformation but an addition. The instruments continue their primary function; the debris tracking is a data analysis application layered onto the existing data stream. That kind of infrastructure reuse is exactly the approach that makes debris monitoring economically tractable alongside the much larger investments in debris cleanup operations at orbital altitude.

Official Sources