How Anti-Satellite Weapons Actually Work: Kinetic Kill, Co-Orbital Attack, and the Debris Problem

Space is not a sanctuary. Every military satellite in orbit follows a predictable trajectory, broadcasts radio-frequency energy toward the ground, and carries no armour plating. A satellite in low Earth orbit (LEO) at 500 kilometres altitude completes one revolution every 94 minutes, and its orbital elements can be determined to within a few metres using ground-based radar. Once you know where a satellite will be and when it will be there, destroying it becomes an engineering problem rather than a physics mystery. The engineering, however, has consequences that persist for centuries.

This article covers the four broad categories of anti-satellite (ASAT) weapons: direct-ascent kinetic interceptors, co-orbital attack vehicles, directed-energy systems, and electronic or cyber attacks against space systems. It then addresses the debris problem that kinetic ASAT tests create, the space situational awareness infrastructure that tracks debris, the tentative steps toward debris mitigation, how satellite operators are redesigning constellations for survivability, and the international legal landscape that has so far failed to prevent any of this.

Why Satellites Are Vulnerable

A satellite in orbit is one of the most predictable objects in existence. Newtonian mechanics dictates its path; perturbations from atmospheric drag, solar radiation pressure, gravitational harmonics (J2 oblateness, lunar and solar third-body effects) are well-modelled. The US Space Surveillance Network (SSN) maintains a catalogue of roughly 30,000 tracked objects larger than 10 centimetres in LEO, and the orbital state vectors for operational satellites are publicly available in the form of Two-Line Element (TLE) sets. Propagating a TLE forward using SGP4/SDP4 gives positional accuracy on the order of one to two kilometres over 24 hours. Military tracking systems, using dedicated radar and electro-optical sensors, do considerably better.

Satellites are built to survive the thermal and radiation environment of space, not to withstand physical impact. A typical LEO Earth-observation satellite has a mass of perhaps 500 to 3,000 kilograms, a structural bus made of aluminium honeycomb or carbon-fibre composite panels, solar arrays spanning several metres, and no redundant shielding beyond micrometeoroid protection (usually Whipple shields or multi-layer insulation rated for particles smaller than one centimetre). A collision at orbital velocities transfers enormous kinetic energy. At a closing speed of 10 km/s, a one-kilogramme impactor delivers 50 megajoules of energy, roughly equivalent to 12 kilograms of TNT. No satellite is designed to survive that.

The predictability of orbits, the fragility of structures, and the reliance on ground-station communications create three distinct attack surfaces: the satellite body itself, the communication links (uplink and downlink), and the ground segment. Different ASAT weapon categories exploit different surfaces.

Direct-Ascent Anti-Satellite Weapons

Direct-ascent ASAT (DA-ASAT) weapons are the conceptually simplest approach: launch a missile from the ground (or from an aircraft), guide it to intercept a satellite, and destroy the target through kinetic impact. No explosive warhead is required. The closing velocity supplies more than enough energy.

The Physics of Kinetic Kill

Consider an interceptor launched from mid-latitudes against a satellite in a 500-kilometre polar orbit. The interceptor ascends on a ballistic trajectory, reaching apogee near the target's altitude. If the satellite is moving at approximately 7.6 km/s and the interceptor arrives with a residual velocity of 2 to 3 km/s, the closing velocity depends on geometry. A head-on intercept can produce closing speeds of 10 to 14 km/s. The kinetic energy at impact scales as one-half times mass times velocity squared, so even a small kill vehicle (10 to 20 kilograms) delivers hundreds of megajoules.

The interceptor does not need to carry explosives. It needs only a seeker (infrared, visible-band, or radar), a divert-and-attitude-control system (DACS) for terminal guidance, and enough structural mass to survive until impact. This is the same "hit-to-kill" technology used in ballistic missile defence interceptors like the SM-3 and Ground-Based Interceptor (GBI). In fact, several DA-ASAT systems are derivatives of or share technology with missile defence programmes.

ASM-135 ASAT (United States, 1985)

The first successful Western DA-ASAT test used a modified two-stage SRAM missile launched from an F-15 Eagle flying at high altitude. On 13 September 1985, Major Wilbert "Doug" Pearson launched the ASM-135 against P78-1 Solwind, a US scientific satellite in a 525-kilometre orbit. The Miniature Homing Vehicle (MHV) kill vehicle used an infrared seeker and eight small rocket motors arranged in a ring for terminal manoeuvring. It struck Solwind at a closing velocity of approximately 14 km/s, creating at least 285 catalogued debris fragments. Most of the debris re-entered within a decade due to the relatively low orbit, but the test demonstrated the operational concept.

The programme was cancelled after this single intercept test, partly due to congressional opposition and partly because the debris implications were already apparent. The US did not conduct another kinetic ASAT test for over two decades.

Operation Burnt Frost (United States, 2008)

In February 2008, the US Navy launched a modified SM-3 Block IA missile from the cruiser USS Lake Erie to intercept USA-193 (NROL-21), a malfunctioning National Reconnaissance Office satellite carrying approximately 450 kilograms of frozen hydrazine propellant. The satellite was in a decaying orbit at roughly 247 kilometres altitude, and the stated justification was to prevent the hydrazine tank from surviving re-entry and reaching the ground intact.

The SM-3 struck USA-193 at a closing velocity of approximately 9.8 km/s. Because the intercept occurred at very low altitude (just above the atmosphere), the resulting debris re-entered within weeks. This was a deliberate choice: destroying a satellite at 247 kilometres minimises the long-term debris problem, because atmospheric drag rapidly clears fragments at that altitude. The Pentagon described the operation as a public safety measure, but it also served as a de facto demonstration that the Aegis Ballistic Missile Defence system could function as a DA-ASAT weapon with minimal software modifications.

SC-19 and the Fengyun-1C Test (China, 2007)

On 11 January 2007, China launched a ground-based SC-19 interceptor (believed to be derived from the DF-21 medium-range ballistic missile with a kinetic kill vehicle upper stage) against Fengyun-1C (FY-1C), a defunct Chinese weather satellite in an 865-kilometre sun-synchronous orbit. The intercept succeeded, and the satellite was completely fragmented.

This single test produced over 3,500 pieces of trackable debris (larger than 10 centimetres) and an estimated 32,000 fragments larger than one centimetre. As of early 2026, roughly 3,000 of the catalogued FY-1C fragments remain in orbit. At 865 kilometres, the atmospheric density is so low that orbital decay takes decades to centuries. Statistical modelling by ESA and NASA indicates that many FY-1C fragments will remain in orbit for 100 to 300 years.

The FY-1C test increased the tracked debris population in LEO by approximately 25% overnight. It remains the single most polluting event in the history of spaceflight, and it demonstrated with brutal clarity why kinetic ASAT tests at high altitude are catastrophically irresponsible.

Mission Shakti (India, 2019)

On 27 March 2019, India's Defence Research and Development Organisation (DRDO) launched a modified PDV Mark II interceptor against Microsat-R, an Indian test satellite in a 283-kilometre orbit. The test, designated Mission Shakti, was designed to minimise debris: the low altitude meant that atmospheric drag would clear fragments relatively quickly.

The intercept produced approximately 400 trackable debris pieces. Most re-entered within months, though some fragments were briefly tracked at altitudes above 1,000 kilometres (boosted by the collision energy), raising concerns at the International Space Station (ISS), which orbits at approximately 420 kilometres. India explicitly cited the low-altitude approach as responsible ASAT testing, drawing a contrast with the Chinese test at 865 kilometres.

PL-19/DN-3 Nudol and the Kosmos-1408 Test (Russia, 2021)

On 15 November 2021, Russia launched a PL-19 Nudol interceptor (a direct-ascent missile originally developed for Moscow's A-235 ballistic missile defence system) against Kosmos-1408, a defunct Soviet ELINT (electronic intelligence) satellite launched in 1982. Kosmos-1408 was in a 487-kilometre orbit with a mass of approximately 2,200 kilograms.

The intercept created over 1,500 trackable debris fragments and an estimated 10,000 to 15,000 pieces larger than one centimetre. The ISS crew (including two Russian cosmonauts) were ordered to shelter in their Crew Dragon and Soyuz vehicles as the debris cloud passed nearby. Several fragments were tracked at altitudes overlapping with the ISS orbit, creating a persistent conjunction risk.

At 487 kilometres, the debris will persist for years to decades. The test was widely condemned, including by ESA, NASA, and the European Union. It provided additional impetus for the United States to declare a unilateral moratorium on destructive DA-ASAT testing in April 2022.

Co-Orbital Anti-Satellite Systems

Co-orbital ASAT (CO-ASAT) weapons take a different approach. Instead of a ground-launched missile on a brief ballistic trajectory, a co-orbital system places a vehicle into orbit near the target satellite, manoeuvres into proximity, and then attacks (or inspects, or interferes with) the target from close range. This category is harder to detect, harder to attribute, and potentially harder to defend against.

Rendezvous and Proximity Operations

The physics of co-orbital approach involves Hohmann transfer orbits, phasing manoeuvres, and relative-motion dynamics described by the Clohessy-Wiltshire (Hill's) equations. A spacecraft in a slightly lower orbit moves faster than the target; by raising its orbit at the right moment, it can approach the target from below and behind. The final approach phase uses thrusters for stationkeeping, typically within a few kilometres to a few hundred metres.

Rendezvous and proximity operations (RPO) are the same techniques used for docking with the ISS or servicing missions. The difference between a cooperative and a non-cooperative RPO is that the target does not provide a docking port, transponder beacon, or fiducial markers. The approaching vehicle must rely on its own sensors (lidar, visible-band cameras, infrared imagers) to determine range, bearing, and relative velocity.

RPO requires relatively modest delta-v. A satellite already in a similar orbital plane to the target might need only tens to hundreds of metres per second of delta-v for phasing and approach manoeuvres. This is well within the capability of a small spacecraft with a few hundred kilograms of hydrazine or a compact electric propulsion system.

Russia's Co-Orbital Programme

Russia has been the most active operator of suspected co-orbital ASAT or inspector vehicles in recent years. Several cases illustrate the pattern.

Kosmos 2542 and 2543 (2019-2020): In November 2019, Russia launched Kosmos 2542 into an orbit that placed it near USA-245 (a classified National Reconnaissance Office satellite believed to be a KH-11 electro-optical reconnaissance platform). In December 2019, Kosmos 2542 released a sub-satellite, designated Kosmos 2543, which manoeuvred independently. In July 2020, Kosmos 2543 itself released a small projectile at high relative velocity, which the US Space Command characterised as an in-orbit weapons test. The sub-satellite-deploying-sub-satellite architecture suggests a Russian interest in co-orbital inspection and potential kinetic attack.

Luch/Olymp (2014-present): Russia's Luch (also called Olymp) satellite was launched in September 2014 into geostationary orbit (GEO). Over the following years, it manoeuvred repeatedly to park itself near Western military and commercial communications satellites, including Intelsat platforms and French-Italian Syracuse military communications satellites. France publicly protested in 2018 after Luch positioned itself between two French military satellites. While Luch may be conducting signals intelligence rather than physical attack preparation, the ability to manoeuvre close to GEO assets demonstrates a co-orbital capability that could be weaponised.

Kill Mechanisms for Co-Orbital Weapons

A co-orbital vehicle in proximity to a target satellite has several attack options:

Kinetic impact: Simply manoeuvring into the target. Even at low relative velocity (a few metres per second), the collision can be destructive. At higher closing speeds (tens to hundreds of metres per second achievable with a final thrust burn), the satellite is certainly destroyed.

Robotic arm or grapple: Seizing the target and altering its orbit, de-spinning it, or physically breaking components. This requires a sophisticated robotic system but leaves the target intact (if the goal is capture rather than destruction). A multi-jointed robotic arm with force-torque sensing at the end effector can grip antenna booms, solar array yokes, or structural hard points on the target bus. Once grappled, the attacker can apply a sustained thrust to de-orbit the target into the atmosphere, push it into a useless graveyard orbit, or simply torque the bus until structural joints fail. The technical heritage for this comes directly from civil robotic servicing programmes: Canada's Canadarm2 on the ISS demonstrated autonomous grapple of free-flying objects, and DARPA's Robotic Servicing of Geosynchronous Satellites (RSGS) programme (now operated by Northrop Grumman under the name Mission Robotics Vehicle) developed arms specifically designed for gripping non-cooperative spacecraft. Converting a servicing arm into a weapon requires no additional hardware, only different software and intent.

Pellet release or fragmentation device: Releasing a cloud of small projectiles in the target's orbital path. Even tiny objects at relative velocities of hundreds of metres per second can destroy solar arrays, antennas, and optical systems. The co-orbital vehicle does not need to collide with the target; it only needs to position itself ahead of the target in the same orbit and release a dispersal charge. A shaped charge that scatters a few hundred tungsten or steel pellets, each a few millimetres in diameter, across the target's orbital path creates a shotgun effect. At closing velocities of even 50 to 100 metres per second (achievable with a small separation manoeuvre), each pellet carries enough energy to puncture thermal blankets, sever wiring harnesses, and crater optical surfaces. The target satellite would appear to suffer a sudden, catastrophic multi-point failure with no obvious external cause unless debris tracking systems happened to observe the pellet release.

Electronic interference at close range: Placing a jammer or spoofing device within metres of the target satellite's antennas, overwhelming the legitimate signals with minimal power.

Spray or coating: Applying a chemical or paint to the target's optics or solar arrays, blinding cameras or reducing power generation. This sounds exotic but requires remarkably little technology: a small canister of opaque liquid and a cold-gas thruster for approach. A pressurised reservoir containing an opaque, fast-curing polymer (similar to commercially available conformal coatings) and a nozzle assembly could deposit a thin film over an optical aperture from a distance of a few metres. In microgravity, the liquid would travel in a coherent stream or fine mist without dripping, making targeting easier than it would be on Earth. Even a partial coating on a primary mirror or lens element would degrade image quality enough to render a reconnaissance satellite operationally useless. Against solar arrays, a coating that absorbs sunlight without converting it to electricity would cause localised heating and power loss. Because no debris is created and no kinetic impact occurs, the attack would be nearly impossible to detect from ground-based SSA sensors; only the target operator, noticing sudden and unexplained degradation, would know something had happened.

The co-orbital approach has a significant advantage for the attacker: attribution is difficult. A satellite that stops functioning might have suffered a technical failure, and unless the proximity of an inspector vehicle was observed and recorded, proving that an attack occurred is challenging. This ambiguity is strategically valuable.

Directed-Energy Anti-Satellite Weapons

Directed-energy weapons (DEW) for ASAT applications include ground-based lasers, space-based lasers, and high-power microwave systems. Of these, ground-based laser systems are the most mature, though their capabilities remain limited compared to kinetic options.

Laser Dazzling and Blinding

The simplest directed-energy ASAT technique is laser dazzling: illuminating a satellite's electro-optical sensor with a laser beam intense enough to saturate or temporarily blind the detector, but not powerful enough to cause permanent physical damage. This is a reversible, non-destructive form of interference.

Dazzling an imaging satellite requires placing a laser spot on the satellite's aperture. For a satellite at 500 kilometres altitude, this requires a beam with a divergence small enough to maintain useful irradiance at that range. The diffraction-limited beam divergence for a laser with wavelength lambda and aperture diameter D is approximately 1.22 times lambda divided by D. For a 1-micrometre wavelength laser with a 1-metre transmitting telescope, the diffraction-limited divergence is about 1.22 microradians, producing a spot diameter of roughly 0.6 metres at 500 kilometres. In practice, atmospheric turbulence degrades this considerably.

Atmospheric Propagation and the Fried Parameter

The primary obstacle for ground-based laser ASAT systems is atmospheric turbulence. The Fried parameter (r0) characterises the coherence length of the atmosphere: the diameter over which the wavefront remains approximately flat. At visible and near-infrared wavelengths, r0 is typically 5 to 20 centimetres at sea level for zenith viewing, scaling with wavelength as lambda to the power of 6/5. For a 1-micrometre laser at a good astronomical site (Mauna Kea, Paranal), r0 might be 15 to 20 centimetres.

When the transmitting aperture is much larger than r0, the beam breaks up into speckles and the effective beam divergence degrades to approximately lambda divided by r0 rather than lambda divided by D. A 1-metre telescope with r0 of 15 centimetres transmits a beam that diverges roughly seven times more than the diffraction limit, producing a spot 4 to 5 metres in diameter at 500 kilometres rather than 0.6 metres. The irradiance drops by the square of this ratio, a factor of roughly 50.

Adaptive optics (AO) can partially compensate for atmospheric turbulence. Using a deformable mirror with hundreds to thousands of actuators, driven by wavefront sensor measurements of a guide star or the target itself, AO systems can recover much of the diffraction-limited performance. Military AO systems at facilities like the Starfire Optical Range in New Mexico have demonstrated near-diffraction-limited beam propagation through the atmosphere. However, AO correction degrades for targets far from the guide star (the isoplanatic angle is typically a few arcseconds at visible wavelengths) and becomes increasingly difficult for large zenith angles where the atmospheric path length increases.

Power Requirements for Physical Damage

Dazzling a satellite sensor requires relatively modest power: a few watts to kilowatts of laser power in a well-collimated beam can saturate a CCD or CMOS detector at 500 kilometres range. Permanent blinding (destroying the focal plane array through thermal damage) requires more, perhaps tens of kilowatts delivered on target.

Physically damaging a satellite's structure, burning through aluminium panels, severing cables, or rupturing propellant tanks, requires a completely different scale of power. The irradiance needed to melt or ablate aluminium at a useful rate is on the order of 10 to 100 kilowatts per square centimetre. At 500 kilometres range, even with perfect adaptive optics, delivering this irradiance with a ground-based laser requires a system in the megawatt class with a transmitting aperture of several metres.

No publicly acknowledged ground-based laser system has demonstrated the ability to physically destroy a satellite at orbital distances. The US, Russia, and China are all believed to have programmes aimed at satellite dazzling and blinding. China reportedly illuminated a US satellite with a ground-based laser in 2006, though the details remain classified.

Space-Based Laser Concepts

Placing a laser weapon in orbit eliminates the atmospheric propagation problem but introduces the challenges of generating and dissipating large amounts of power in a vacuum. A space-based laser needs a power source (solar arrays or a nuclear reactor), a beam director, a thermal management system, and the laser itself. The mass budget for a space-based laser capable of physically damaging satellites at useful ranges (hundreds of kilometres) would be enormous, likely tens of thousands of kilograms. No country is known to have deployed such a system, though the Soviet Union pursued the Skif programme in the 1980s (the Polyus spacecraft launched in 1987 was reportedly a prototype orbital weapons platform that failed to reach orbit due to a guidance error during insertion).

The thermal management problem alone makes space-based lasers extremely difficult. A laser operating at 20% electrical-to-optical efficiency must dissipate four watts of waste heat for every watt of beam power. In space, the only heat rejection mechanism is thermal radiation (no convective or conductive cooling is available), which requires enormous radiator panels. A one-megawatt laser generating four megawatts of waste heat would need radiator area on the order of hundreds of square metres, depending on operating temperature. This creates a spacecraft with the mass and size of a large space station, an obvious and vulnerable target in its own right.

High-power microwave (HPM) weapons represent a different directed-energy approach for ASAT applications. Rather than heating a target to destruction, an HPM system radiates a broad beam of microwave energy that couples into the target satellite's electronics through antennas, cable runs, and apertures in the shielding. The energy induces currents in circuit traces and semiconductor junctions, causing upset (temporary malfunction), latchup (requiring a power cycle to recover), or burnout (permanent destruction of components). HPM is particularly effective against satellites because spacecraft electronics are designed for the relatively benign electromagnetic environment of orbit, not for resistance to intentional microwave illumination. A ground-based or space-based HPM source would not need to deliver nearly as much power as a structural-damage laser, because the damage mechanism is electronic disruption rather than thermal ablation. The US Air Force Research Laboratory has investigated HPM for counter-electronics applications for decades, and the physics applies equally to terrestrial and orbital targets.

Electronic and Cyber Attacks on Space Systems

Not all ASAT weapons are kinetic or directed-energy. Satellites depend on radio-frequency links for commanding (uplink), telemetry (downlink), and data relay. They depend on ground stations for orbit determination, mission planning, and data processing. Both the RF links and the ground infrastructure are vulnerable to electronic and cyber attack.

Uplink Jamming and Spoofing

A satellite receives commands on a specific frequency through its command receiver. If an adversary can transmit a stronger signal on the same frequency, the legitimate commands are drowned out. This is uplink jamming. If the adversary can mimic the command protocol (which may or may not be encrypted), they can inject false commands; this is command spoofing.

Military satellites generally use encrypted, frequency-hopping, or spread-spectrum command links that are resistant to simple jamming. Commercial satellites, particularly older ones, may use simpler protocols. GPS satellites transmit on known frequencies, and the civil GPS signal (L1 C/A) has no encryption, making it vulnerable to both jamming and spoofing. Military GPS (M-code, P(Y)-code) uses encrypted signals with significantly better anti-jam characteristics.

The Viasat KA-SAT Cyber Attack (February 2022)

The most significant publicly known cyber attack against a space system occurred on 24 February 2022, coinciding with the Russian invasion of Ukraine. Attackers (later attributed to Russian military intelligence by the EU, UK, and US) targeted the KA-SAT network operated by Viasat. The attack exploited a misconfigured VPN appliance in the ground segment, gained access to the network management system, and pushed a destructive firmware update to tens of thousands of Surfbeam2 satellite modems across Europe.

The attack bricked approximately 30,000 modems in Ukraine and neighbouring countries. It disrupted communications for Ukrainian military units, government agencies, and civilians. Collateral damage extended well beyond Ukraine: 5,800 Enercon wind turbines in Germany lost their remote monitoring and control links (the turbines continued to generate power but could not be remotely managed). Modems in France, Italy, Poland, the Czech Republic, and other EU member states were also affected.

The Viasat attack demonstrated several important points. First, the ground segment is often the weakest link in a space system. The satellite itself was unaffected; the attack targeted the terrestrial network management infrastructure. Second, cyber attacks on space systems can have wide-area effects that cross national borders. Third, commercial space infrastructure is deeply integrated with both military and civilian critical infrastructure, making it a high-value target.

The Kessler Syndrome and the Debris Cascade Problem

In 1978, NASA scientist Donald Kessler published a paper describing a scenario in which the density of objects in LEO becomes high enough that collisions between them produce more debris than natural decay processes remove. Once this threshold is crossed, a self-sustaining cascade of collisions occurs, progressively filling orbital altitude bands with fragments that render them unusable for centuries.

Collision Physics

Two objects in crossing orbits in LEO can collide at velocities ranging from nearly zero (for objects in similar orbits) to approximately 14 km/s (for objects in opposed polar orbits). The average collision velocity in LEO is approximately 10 km/s. At these speeds, the collision is hypervelocity: the impact duration is microseconds, the pressures exceed the material strength by orders of magnitude, and both objects are catastrophically fragmented. The result is not two pieces; it is thousands of fragments spanning a wide range of sizes.

The NASA standard breakup model (EVOLVE/ORDEM) predicts the number and size distribution of fragments from a catastrophic collision. For two objects with a combined mass of 2,000 kilograms colliding at 10 km/s, the model predicts thousands of fragments larger than one centimetre, hundreds larger than 10 centimetres, and millions of sub-centimetre particles. Each of these fragments becomes a potential impactor for other satellites.

Collision Probability

The probability that a given satellite collides with a catalogued debris object over a given time period depends on the spatial density of debris in the relevant orbital regime, the collision cross-section of the satellite, and the relative velocity. For a satellite with a cross-section of 10 square metres in an orbit with a spatial density of 10^-8 objects per cubic kilometre (roughly the current density near 800 kilometres altitude), the annual collision probability with a catalogued object is on the order of 10^-4 to 10^-3. This sounds small, but for a constellation of 1,000 satellites operating for 10 years, the expected number of collisions approaches one.

The more insidious problem is the sub-catalogue population: objects between 1 and 10 centimetres that are too small to track reliably but large enough to destroy a satellite. The population of these objects is estimated at 500,000 to 1,000,000 in LEO. A one-centimetre aluminium sphere at 10 km/s delivers approximately 50 kilojoules, enough to penetrate any spacecraft wall and cause catastrophic damage to subsystems.

Current State of the Debris Environment

As of early 2026, the US SSN tracks approximately 30,000 objects larger than 10 centimetres in Earth orbit. The ESA MASTER model estimates the total population as approximately 36,000 objects larger than 10 centimetres, 1,000,000 objects between 1 and 10 centimetres, and 130,000,000 objects between 1 millimetre and 1 centimetre.

The debris population is concentrated in certain altitude bands. The most congested region is 750 to 900 kilometres, where the Fengyun-1C debris, the Cosmos-Iridium collision debris (from the 2009 collision between Cosmos 2251 and Iridium 33), and a large number of spent rocket bodies and defunct satellites accumulate. This altitude band is above the region where atmospheric drag provides significant natural debris removal (drag is negligible above roughly 700 kilometres for objects with typical area-to-mass ratios).

The 2009 Cosmos-Iridium collision was the first accidental catastrophic collision between two catalogued objects. Cosmos 2251 (a defunct Russian military communications satellite, approximately 900 kilograms, launched in 1993 and decommissioned in 1995) collided with Iridium 33 (an operational US communications satellite, approximately 560 kilograms) at a relative velocity of 11.7 km/s at an altitude of 789 kilometres. The collision produced over 2,300 trackable fragments, most of which will remain in orbit for decades. The collision was notable not only for its severity but for its preventability. Iridium 33 was operational and manoeuvrable; Cosmos 2251 was not. However, the conjunction warning systems at the time did not flag the event with sufficient confidence or lead time for Iridium to execute an avoidance manoeuvre. The US Joint Space Operations Center (JSpOC) issued thousands of conjunction warnings per day, and the probability estimate for this particular close approach did not rise above the threshold that would have triggered a manoeuvre. After the collision, Iridium Communications and the US military significantly revised their conjunction screening procedures, lowering the probability threshold for warnings and improving the timeliness of conjunction data messages. The collision debris clouds from Cosmos 2251 and Iridium 33 spread into distinct orbital shells over the following months, with fragments tracked at altitudes ranging from 200 to over 1,300 kilometres. As of 2026, approximately 1,800 of the original 2,300 tracked fragments remain in orbit.

Modelling the Cascade

ESA, NASA, and several university groups run long-term debris environment models (ESA's MASTER/DELTA, NASA's LEGEND/ORDEM, University of Southampton's DAMAGE). These models simulate the evolution of the debris population over 100 to 200 year timescales, accounting for new launches, explosions, collisions, atmospheric drag, and debris mitigation measures.

The models consistently show that even with no new launches (a plainly unrealistic assumption), the debris population at 800 to 1,000 kilometres will continue to grow due to collisions among existing objects. NASA's LEGEND model has run Monte Carlo simulations with 200 or more randomised scenarios, projecting the debris environment forward to the year 2200. In the "no new launches" baseline, the model predicts approximately four to five catastrophic collisions per decade in the 700 to 1,000 kilometre band, each producing hundreds of new trackable fragments that seed further collisions. The ESA MASTER/DELTA suite produces similar results, with their reference scenario (based on current launch rates and 60% compliance with the 25-year de-orbit guideline) predicting a 30 to 40% increase in the total catalogued population by 2100, driven primarily by collision-generated fragments rather than new launches.

A particularly concerning scenario modelled by both NASA and the University of Southampton's DAMAGE model involves the large intact objects (spent rocket bodies and defunct satellites) that currently populate the 800 to 1,000 kilometre band. Approximately 2,000 objects with masses exceeding 500 kilograms orbit in this region, and each collision between two such objects can produce 1,000 or more trackable fragments. The models suggest that a single collision between two large rocket bodies in this band would increase the collision probability for every other object at that altitude by a measurable amount, potentially triggering a follow-on collision within years rather than decades.

With realistic launch rates and current debris mitigation compliance, the models predict a slow but steady increase in the collision rate, with the cascade threshold potentially being crossed within 50 to 100 years in the most congested altitude bands.

The critical finding is that removing five to ten large objects per year from the most congested orbits could stabilise the debris environment and prevent the cascade. This provides the motivation for active debris removal (ADR) missions.

Space Situational Awareness

Tracking objects in orbit, predicting conjunctions, and providing collision avoidance warnings is the domain of space situational awareness (SSA), also called space domain awareness (SDA). Several sensor networks contribute to this mission.

US Space Surveillance Network

The US SSN is the most capable SSA system, operated by US Space Command (formerly USSTRATCOM). It consists of approximately 30 sensors worldwide, including phased-array radars (AN/FPS-85 in Florida, AN/FPS-132 in the UK, Greenland, Alaska, and California), mechanical tracking radars, and electro-optical telescopes (Ground-based Electro-Optical Deep Space Surveillance, or GEODSS, sites in Hawaii, Diego Garcia, and New Mexico).

The most significant recent addition to the SSN is the Space Fence, an S-band phased-array radar on Kwajalein Atoll in the Marshall Islands, which achieved initial operational capability in 2020. The Space Fence can detect objects as small as 5 centimetres in LEO and has dramatically increased the number of tracked objects. It represents a generational improvement in SSA capability.

European Space Surveillance and Tracking

The EU established the Space Surveillance and Tracking (EU SST) programme in 2015, now part of the EU Space Programme for Space Surveillance and Tracking (EUSST). The contributing sensor network includes:

GRAVES (Grand Reseau Adapte a la Veille Spatiale): A French bistatic radar system near Dijon, with the transmitter at Broye-les-Pesmes and receivers at Apt. GRAVES operates in the VHF band (143.05 MHz) and can detect objects in LEO down to approximately 50 centimetres. It provides autonomous European LEO surveillance capability independent of US data.

TIRA (Tracking and Imaging Radar): Operated by the Fraunhofer Institute for High Frequency Physics and Radar Techniques (FHR) near Bonn, Germany. TIRA is an L-band tracking radar with a 34-metre dish that can track objects as small as 2 centimetres in LEO and produce radar images of satellites at ranges up to 1,000 kilometres. It is one of the most capable single-dish tracking radars in Europe.

Spanish S3T radar: A surveillance radar developed by Indra for the Spanish Space Surveillance Operations Centre, contributing LEO surveillance data.

Electro-optical sensors: Multiple telescopes across Spain, France, Germany, Poland, and other EU member states provide optical tracking of objects in MEO and GEO, where radar performance degrades due to the inverse-fourth-power range dependence.

The EUSST consortium issues conjunction data messages (CDMs) to European satellite operators and coordinates collision avoidance manoeuvre planning. It also provides re-entry predictions for uncontrolled objects.

ESA Space Safety Programme

ESA operates its own Space Debris Office at ESOC in Darmstadt, Germany, and maintains the DISCOS database of known space objects. ESA's Space Safety programme encompasses SSA, planetary defence (asteroid deflection), and space weather monitoring. ESA has invested in developing next-generation SSA sensors, including the proposed Flyeye telescopes (wide-field optical survey telescopes for NEO and debris detection) and has contributed to the European Southern Observatory's (ESO) laser guide star adaptive optics technology, which has applications for debris tracking and characterisation.

Debris Mitigation and Active Removal

The IADC 25-Year Rule

The Inter-Agency Space Debris Coordination Committee (IADC), comprising the space agencies of 13 countries plus ESA, adopted guidelines recommending that satellites in LEO be de-orbited or moved to a disposal orbit within 25 years of end of mission. This guideline was endorsed by the UN Committee on the Peaceful Uses of Outer Space (COPUOS) in 2007.

Compliance with the 25-year rule has improved over time but remains incomplete. As of 2025, approximately 60 to 70% of missions ending in LEO comply with the guideline. The remaining 30 to 40%, including failed satellites, spent rocket upper stages, and operators who choose not to expend fuel on de-orbit burns, contribute to the long-term debris growth. The FCC in the United States adopted a stricter 5-year de-orbit rule in September 2022, and ESA has proposed a "zero debris" policy for its own missions by 2030.

Active Debris Removal

Recognising that mitigation alone cannot stabilise the debris environment, several organisations are developing active debris removal (ADR) technologies.

ClearSpace-1 (ESA): ESA contracted ClearSpace SA (a Swiss startup spun out of EPFL) in 2020 to conduct the first ADR demonstration mission. ClearSpace-1 will use a capture vehicle with four robotic arms to grab VESPA (Vega Secondary Payload Adapter), a derelict ESA-owned object in a 660 by 800 kilometre orbit with a mass of approximately 112 kilograms. After capture, the combined stack will perform a controlled de-orbit, burning up in the atmosphere. The mission, originally planned for 2025, has been delayed to 2027 following an accidental debris impact on VESPA in August 2023 that fragmented part of the target, complicating the capture geometry.

Astroscale ELSA-d (Japan/UK): Astroscale, a Japanese-founded company with offices in Harwell, UK, launched the End-of-Life Services by Astroscale demonstration (ELSA-d) in March 2021. The mission consisted of a servicer spacecraft and a client spacecraft (simulating a debris object) equipped with a ferromagnetic docking plate. The servicer demonstrated repeated capture and release of the client using a magnetic docking mechanism. ELSA-d validated the capture technology but targeted only cooperative (plate-equipped) clients. Astroscale's follow-on ADRAS-J mission, launched in February 2024, performed the first RPO with an actual piece of debris (the upper stage of a Japanese H-2A rocket) without a docking plate, using only lidar and cameras for proximity navigation.

Other concepts: Various other ADR technologies have been proposed and are in development, including net capture (RemoveDEBRIS, demonstrated in orbit in 2018), harpoon capture (also demonstrated on RemoveDEBRIS), laser ablation (using a ground or space-based laser to ablate material from a debris object's surface, creating a small thrust to lower its orbit), electrodynamic tethers (deploying a long conducting tether from a debris object that interacts with Earth's magnetic field to generate drag), and drag augmentation devices (inflatable structures that increase a debris object's cross-section and atmospheric drag).

The challenge for all ADR concepts is economics. Removing one object costs tens of millions of euros per mission. Removing five to ten large objects per year (the minimum rate needed to stabilise the environment) implies an annual cost of hundreds of millions of euros, with no immediate commercial return. This is a classic tragedy-of-the-commons problem: the cost of debris removal is borne by the remover, while the benefit (a cleaner orbital environment) is shared by all operators.

Resilient Space Architectures

The vulnerability of large, expensive, small-number satellite constellations to ASAT attack has driven a rethink of space architecture, particularly for military systems.

Disaggregation

Traditional military space architecture concentrates many capabilities on a single, large, exquisite satellite. The US Space-Based Infrared System (SBIRS) satellites, for example, each provide missile warning, missile defence cueing, battlespace characterisation, and technical intelligence from a single GEO platform costing over 5 billion euros. Destroying one such satellite removes an enormous amount of capability.

Disaggregation spreads these functions across multiple, smaller, cheaper satellites. Instead of one satellite doing everything, ten satellites each do one thing. Destroying one removes only a fraction of the overall capability, and the adversary must expend ten ASAT weapons instead of one.

Proliferated LEO Constellations

The Space Development Agency (SDA), established by the US Department of Defense in 2019 (now part of the US Space Force), is building the Proliferated Warfighter Space Architecture (PWSA): a mesh network of hundreds of small satellites in LEO providing missile tracking, data relay, and custody transfer. The first tranche of PWSA satellites (Tranche 0) launched in 2023 and 2024, with Tranche 1 (approximately 150 satellites) in deployment as of early 2026.

The logic is straightforward: a constellation of 200 satellites in LEO, each costing perhaps 10 to 20 million euros, is far harder to neutralise than two satellites in GEO costing 5 billion euros each. The adversary would need to expend 200 ASAT weapons, each of which costs tens of millions of euros and creates a debris cloud. The cost exchange ratio strongly favours the defender.

European nations are pursuing similar concepts. The IRIS2 (Infrastructure for Resilience, Interconnectivity, and Security by Satellite) programme, approved by the EU in 2023, will provide a sovereign European multi-orbit communications constellation with a security component. The architecture deliberately spans LEO, MEO, and GEO orbits, so that an ASAT attack against one orbital regime does not eliminate the entire system. France has been particularly active in rethinking space resilience following its experience with Russian Luch/Olymp manoeuvres near its Syracuse military communications satellites. The French Ministry of Armed Forces published a space defence strategy in 2019 that explicitly addressed ASAT threats and called for active defence measures, including the ability to "characterise threats" and "respond" in the space domain. France's next-generation CERES signals intelligence constellation and the CSO (Composante Spatiale Optique) optical reconnaissance system were designed with resilience considerations, including multiple satellites per constellation and hardened command links. Germany's SatcomBw Stage 3 programme similarly incorporates redundancy and secure communications for Bundeswehr space assets. The UK's Skynet 6 programme is incorporating disaggregation principles into next-generation military communications, with Skynet 6A scheduled to launch by 2025 and additional capacity planned through commercial augmentation contracts rather than reliance on a single exquisite platform.

Rapid Reconstitution

If satellites can be destroyed, the ability to rapidly replace them becomes a strategic capability. Rapid reconstitution involves maintaining ground spares, pre-positioning satellites in orbit for rapid deployment, and developing responsive launch capabilities that can place replacement satellites in orbit within days to weeks rather than months to years.

The US Space Force Tactically Responsive Space (TacRS) programme, along with developments in commercial launch (SpaceX Falcon 9 achieving 40+ launches per year, Rocket Lab Electron providing dedicated small-satellite launch), provides the foundation for rapid reconstitution. Europe's responsiveness in this area has historically lagged, though ESA's Boost! programme for next-generation launch vehicles and the development of micro-launch capabilities (Isar Aerospace Spectrum, PLD Space Miura 5) aim to close the gap.

International Law and Arms Control

The Outer Space Treaty (1967)

The Treaty on Principles Governing the Activities of States in the Exploration and Use of Outer Space, Including the Moon and Other Celestial Bodies (commonly called the Outer Space Treaty, or OST) is the foundational legal instrument for space activities. It entered into force on 10 October 1967, and as of 2026, 114 states are parties.

The OST prohibits placing nuclear weapons or other weapons of mass destruction in orbit (Article IV). It does not prohibit conventional weapons in orbit, including kinetic ASAT weapons. It does not prohibit ground-based ASAT weapons of any kind. The treaty declares that outer space is free for exploration and use by all states (Article I) and that no state may claim sovereignty over celestial bodies (Article II), but these provisions do not meaningfully constrain ASAT development.

Article IX requires states to conduct their space activities with "due regard" to the interests of other states and to engage in "appropriate international consultations" before proceeding with activities that might cause "potentially harmful interference" with other states' space activities. This provision could arguably apply to ASAT tests that generate debris affecting other nations' satellites, but it has never been successfully invoked for this purpose.

The 2022 US ASAT Test Moratorium

On 18 April 2022, Vice President Kamala Harris announced that the United States would not conduct destructive direct-ascent anti-satellite missile tests and called on other nations to make similar commitments. The moratorium was framed as a norm-setting initiative rather than a legal obligation: it applies only to destructive DA-ASAT tests (those that create debris), not to other forms of ASAT capability (co-orbital, electronic, cyber, directed energy, or non-destructive DA-ASAT tests at very low altitude).

As of early 2026, over 40 states have endorsed the moratorium, including Canada, New Zealand, Japan, South Korea, Germany, France, the United Kingdom, Switzerland, and several other European nations. Neither Russia nor China has endorsed it.

The moratorium is significant but limited. It addresses only the most environmentally destructive form of ASAT testing (kinetic DA-ASAT at medium to high altitude). It does not constrain the development, production, or deployment of ASAT weapons; it constrains only testing. An adversary could develop and field a DA-ASAT weapon using missile defence interceptor technology (which can be tested against non-orbital targets or simulated in software) without ever conducting a debris-generating ASAT test.

UN Open-Ended Working Group

The UN General Assembly established an Open-Ended Working Group (OEWG) on Reducing Space Threats Through Norms, Rules, and Principles of Responsible Behaviours in 2022. The OEWG met through 2022 and 2023 but was unable to reach consensus on a final report due to disagreements between Western states (favouring norms of behaviour) and Russia and China (favouring a treaty banning weapons in space, specifically the PPWT, or Prevention of the Placement of Weapons in Outer Space Treaty, which Western states reject because it does not cover ground-based ASAT weapons and contains no verification provisions).

The fundamental arms-control challenge for ASAT weapons is dual-use technology. A missile defence interceptor is an ASAT weapon. A satellite servicing vehicle is a co-orbital ASAT weapon. A satellite communications jammer is an electronic ASAT weapon. Distinguishing between legitimate and hostile capabilities requires intrusive verification that no state has been willing to accept.

The Strategic Calculus

The proliferation of ASAT capabilities creates a strategic environment with several uncomfortable characteristics.

First, the offence-defence balance strongly favours the attacker. A DA-ASAT missile costs perhaps 10 to 50 million euros. A modern military satellite costs 500 million to 5 billion euros. The cost exchange ratio is 10:1 to 100:1 in favour of the attacker, unless the defender responds with proliferated architectures that change the arithmetic.

Second, kinetic ASAT weapons are indiscriminate in their second-order effects. Destroying a single satellite in a congested orbital regime creates debris that threatens every satellite at similar altitudes, including the attacker's own satellites and those of neutral parties. This is the space equivalent of using nuclear weapons in a confined theatre: the fallout harms everyone.

Third, the threshold for use is ambiguous. Unlike nuclear weapons, which have a clear taboo reinforced by 80 years of non-use, ASAT weapons have been tested destructively at least four times (US 1985, China 2007, India 2019, Russia 2021). There is no established norm against their use in conflict, only a recent and incomplete norm against testing.

Fourth, the dependencies are asymmetric. Nations that rely most heavily on space (the United States and its allies, which use satellite navigation, communications, intelligence, and early warning for virtually all military operations) are the most vulnerable to ASAT attack. Nations that rely less on space have less to lose.

For European nations specifically, the strategic implications are significant. European militaries depend heavily on US GPS (supplemented by Galileo), US and commercial satellite communications, and US intelligence satellites. European sovereign space capabilities are growing (Galileo, Copernicus, IRIS2, Syracuse, SatcomBw, Skynet) but remain limited compared to US assets. An adversary's ASAT campaign against Western space infrastructure would disproportionately affect European forces that lack indigenous alternatives to US space services.

The response to this vulnerability involves the full spectrum of measures discussed above: proliferation of satellites, diversification of orbits, hardening of ground segments, development of European SSA capabilities, investment in rapid reconstitution, and diplomatic efforts to establish norms against destructive ASAT use. Whether these measures will prove sufficient is an open question that will be answered not by engineers but by the strategic choices of nations.