Orbital Debris and Space Traffic

Established Earth orbit is now a managed-traffic environment, and the management is improvised. The European Space Agency’s statistical models put the debris population at more than 36,500 objects larger than 10 cm, roughly one million objects between 1 cm and 10 cm, and on the order of 330 million fragments between 1 mm and 1 cm; surveillance networks maintain custody of a catalogue of roughly 47,000 tracked objects. Any of the centimetre-class objects can destroy a spacecraft, and only the largest class can be tracked.

Established The traffic side has grown faster than the debris side. Active satellites went from about 2,500 in 2020 to more than 10,000 by early 2026, more than half of them SpaceX Starlink spacecraft. SpaceX’s own filings and independent tallies put its collision-avoidance manoeuvres at roughly 300,000 in calendar 2025 — on the order of 800 per day — against about 200,000 in 2024. No other operator publishes comparable data.

Frontier Whether the coordination machinery scales is the open question, and it is institutional as much as technical. Conjunction screening, manoeuvre deconfliction, debris-removal target selection, liability, standards and payment are usually treated as six separate literatures. They are one system: the quality of the environment measurement sets the false-alarm rate of conjunction warnings; the warning rate sets the coordination burden; the coordination burden sets what rules and institutions are needed; and the absence of a payment mechanism decides whether remediation happens at all.

Established This brief joins those pieces. It is deliberately unlike the popular framing: the evidence points not to a cinematic cascade that seals humanity on Earth, but to a slow, measurable degradation of a commons — rising operational cost, rising insurance ambiguity, occasional satellite losses — that every other orbital ambition on this site’s map, from deep space infrastructure to space-based solar power, must transit through.

Established The environment is measured annually and the measurement is credible. ESA’s Space Environment Report, issued each year from the Space Debris Office’s MASTER and DELTA modelling and the public catalogues, is the reference record of the orbital population; its 2025 edition describes launch traffic at record levels, the LEO active population dominated by constellations, and post-mission disposal compliance improving for recently launched payloads while a large stock of legacy rocket bodies and dead satellites remains on orbit for decades. Frontier Population figures below the tracking floor are model outputs, not counts: the one-million figure for 1–10 cm objects carries substantial and honestly stated uncertainty, because it is calibrated on returned-surface impact records and a handful of dedicated radar campaigns.

Established A short list of events created much of the tracked debris. China’s January 2007 Fengyun-1C anti-satellite test produced more than 3,000 tracked fragments and an estimated 32,000 smaller ones, most still aloft. The February 2009 collision of the defunct Cosmos 2251 with the operational Iridium 33 — the first accidental hypervelocity collision of two intact satellites — added roughly 2,000 trackable fragments. Russia’s November 2021 Nudol test against Cosmos 1408 added roughly 1,500 more. In 2024 two further events were widely reported: the breakup of a Long March 6A upper stage in August, generating several hundred tracked fragments in a heavily used sun-synchronous band, and the October breakup of the Intelsat 33e communications satellite in geostationary orbit.

Established Conjunction warning is a public service run by a military unit. The US Space Force’s 18th and 19th Space Defense Squadrons maintain the public catalogue and screen it for close approaches, issuing conjunction data messages (CDMs) to operators worldwide; NASA’s CARA programme performs risk assessment for civil missions. The screening load is enormous: a full run compares on the order of a billion object pairs, several times a day. Dmitry Poisik, the TraCSS programme manager at the US Office of Space Commerce, put the output plainly in 2025: “On a bad day, it’s a million predicted [conjunctions] just in the next week.”

Established The manoeuvre record is dominated by one operator’s unusually transparent filings. SpaceX’s semi-annual constellation status reports to the US Federal Communications Commission are the only sustained public dataset of their kind: roughly 25,000 collision-avoidance manoeuvres in December 2022–May 2023; 50,000 in December 2023–May 2024 (about 275 per day, reported in the 1 July 2024 filing); 50,666 in June–November 2024; 144,404 in December 2024–May 2025; and 148,696 in June–November 2025, split between 44,559 first-generation and 104,137 second-generation satellites. Independent trackers put the calendar-2025 total near 300,000, about fifty percent above 2024.

Frontier The raw counts overstate risk growth, because SpaceX manoeuvres at extraordinarily conservative thresholds — and the sources disagree on the number. The common “actionable” industry threshold is a 1-in-10,000 collision probability; SpaceX’s July 2024 filing cites manoeuvring at 1-in-1,000,000, while 2025 reporting cites 3-in-10,000,000. This brief declares that tie rather than averaging it. Hugh Lewis of the University of Southampton, the most careful outside analyst of the filings, estimated that about half of the December 2023–May 2024 manoeuvres would not have happened under the older threshold, and projects around 80,000 manoeuvres per half-year by 2027; he also notes that rising solar activity has been thinning debris at Starlink altitudes through increased atmospheric drag.

Established The same filings document the failure modes of the warning system itself. In June–November 2024 SpaceX reported three outages of the conjunction-data pipeline lasting 9–24 hours, 93 cases of late-arriving conjunction warnings, and 100 individual satellite outages that prevented risk-mitigation manoeuvres; it also reported that other operators performed fewer than ten coordination manoeuvres with Starlink in the period, and that some operators do not even publish current contact information. The same report records 73 Starlink satellites re-entered and 14 deactivated in six months — disposal at a scale no operator has run before (operator’s own figures throughout).

Established The rulebook is mostly voluntary and recently tightened. The Inter-Agency Space Debris Coordination Committee’s mitigation guidelines (2002, revised 2007) established the 25-year post-mission disposal norm; UN COPUOS adopted parallel guidelines in 2007 and Long-Term Sustainability guidelines in 2019; the Space Safety Coalition published industry best practices in October 2019 with 31 initial endorsers. The binding exception: in September 2022 the US FCC adopted a 5-year deorbit rule for LEO satellites it licenses or admits to the US market, and in October 2023 it issued its first debris enforcement penalty, a $150,000 settlement with DISH over the EchoStar-7 disposal orbit. ESA’s Zero Debris approach commits its own missions to debris-neutrality by 2030, with a Zero Debris Charter signed by well over a hundred entities since late 2023.

Established Civil traffic coordination in the US has been a five-year institutional fight. Space Policy Directive-3 (June 2018) assigned civil space traffic management to the Commerce Department; its Traffic Coordination System for Space (TraCSS) entered beta on 30 September 2024. The FY2026 budget proposal of June 2025 sought to terminate federal TraCSS funding and cut the Office of Space Commerce from $65 million to $10 million, arguing commercial services sufficed; seven industry associations representing over 450 companies objected, and the Space Force made clear it did not want the job back — a retired officer called it an “unfunded mandate,” and an industry executive compared it to asking the Air Force to run every civil control tower. Congress restored $52.5 million for FY2026 (the House committee approved $50 million with “systems integrator” language on 13 May 2026). As of mid-2026 TraCSS served 62 pilot organisations operating more than 11,000 satellites (up from 35 organisations and 10,696 satellites in April), yet remained formally a pilot, with the FY2027 request cutting the office to $11 million and leaving its final institutional form — government-run, contractor-owned, or data-as-a-service — explicitly open.

Frontier Active debris removal has flown reconnaissance, not yet removal. Astroscale’s ELSA-d (2021) demonstrated magnetic capture and release of a prepared client but curtailed later phases after thruster failures. Its ADRAS-J mission, Phase I of JAXA’s Commercial Removal of Debris Demonstration (CRD2) programme, launched in February 2024 and performed the first commercial rendezvous and proximity inspection of a real piece of large debris — a three-tonne H-2A upper stage — closing to reported distances of tens of metres and executing at least one autonomous safe abort (vendor-reported). ESA’s ClearSpace-1, contracted in 2020 at 86 million euros to capture a VESPA payload adapter, was re-scoped in 2024 after its target was itself struck by debris in August 2023, an irony the field noticed; the redesigned mission now aims at the retired PROBA-1 satellite around 2028. Established No large uncooperative object has ever been removed from orbit by anyone.

Frontier What is the carrying capacity of a shell? There is no agreed metric for how much traffic a given altitude band can sustain at a given debris density, manoeuvre threshold and drag regime. Aerospace Corporation researchers, among others, are building orbital-capacity frameworks, but licensing decisions — tens of thousands of satellites approved on paper, with credible projections of 70,000 or more active satellites by 2030 — are being made without one.

Frontier How fast does the debris population compound? Long-run evolutionary models agree on the sign and disagree on the rate. The canonical NASA finding (Liou and Johnson, 2006) is that even with no further launches, collisions among objects already in orbit would keep LEO debris growing in the most crowded bands over century timescales — a slow feedback, not an explosion. Newer stochastic models make sharper claims: one such exercise (KESSYM, circulated via the KeepTrack analysis community) puts critical density thresholds in key shells by 2075 and “irreversible orbital collapse” within 250 years under business-as-usual. Speculative Those specific dates are model artefacts sensitive to launch-rate and solar-cycle assumptions and should be read as scenarios, not forecasts.

Frontier Can conjunction warnings be made decision-grade? A warning is only as good as its covariance. Position uncertainties for catalogued objects are commonly hundreds of metres to kilometres; a single manoeuvre can throw a predicted trajectory off by up to 40 km and corrupt predictions for days unless the operator shares ephemerides. Making false-alarm rates low enough that a manoeuvre means something — without missing the real event — is an open data-fusion problem.

Frontier Which objects should be removed first? Statistical rankings consistently point at large derelict rocket bodies clustered in crowded, high-inertia bands — mass times collision probability times consequence — rather than at the small debris that dominates counts. Researchers such as Darren McKnight have maintained lists of the statistically most concerning derelicts, heavy with Soviet-era upper stages near 800–1,000 km. Removal studies since the early 2010s converge on a striking number: removing roughly five to ten of the right large objects per year stabilises the LEO environment in most models. Nobody removes any.

Frontier What do mass reentries do to the atmosphere? Constellation-scale disposal means tonnes of aluminium and other metals ablating in the mesosphere and stratosphere daily. Measurement campaigns have found spacecraft-derived metals in stratospheric aerosol, and 2024–2025 modelling papers examine alumina accumulation and possible effects on ozone chemistry and atmospheric circulation; magnitudes remain genuinely unresolved. The debris solution — deorbit everything quickly — may have its own environmental bill.

Speculative Can pricing fix the incentive structure? The economics literature treats orbital congestion as a classic open-access externality. Rao, Burgess and Kaffine (PNAS, 2020) computed that an internationally harmonised orbital-use fee, rising to roughly $235,000 per satellite-year by 2040, would maximise the long-run value of the orbital economy. No jurisdiction has tried it, and unilateral fees invite flag-of-convenience licensing.

Established The lethal non-trackable gap. Ground radar reliably catalogues LEO objects down to roughly 5–10 cm. The estimated one million objects between 1 and 10 cm can destroy or disable a satellite yet cannot be tracked or warned against; operators can only shield against the millimetre class and accept the centimetre class as background risk. Closing this gap needs either much larger radar apertures or space-based optical/in-situ sensing at scale, neither funded at the level required.

Frontier Covariance realism and data fusion. Military catalogues, commercial radar networks such as LeoLabs, and operator GPS ephemerides each carry different biases and uncertainty models; fusing them into warnings with honest error bars is unsolved in practice. Classification adds friction: the highest-accuracy tracking data is not shareable.

Established Coordination protocols are pre-industrial. Operator-to-operator deconfliction still runs substantially on email and goodwill. In August 2025 the US government advised operators facing conjunctions with Chinese constellation satellites to email the Beijing Institute of Telecommunications and Tracking Technology. There is no machine-to-machine standard for agreeing who moves, no right-of-way rule, and no obligation to respond.

Frontier Non-cooperative capture. Legacy debris has no grapple fixtures, may tumble, and may shed material when touched. Robotic-arm capture of a tumbling multi-tonne stage has never been done; every current mission (ADRAS-J2, ClearSpace-1) is engineering its first attempt. Capture mechanisms, relative-navigation sensors and fault-tolerant autonomy near a target are all at demonstration maturity.

Established Cost per removal is two orders of magnitude off. Current single-target demonstration missions cost on the order of $100 million against studies calling for five to ten large removals per year indefinitely — independent estimates put a stabilisation campaign at $500 million to $1 billion annually at today’s prices. Handwave The common assertion that “the market will drive costs down as with launch” skips the missing step: launch had paying customers; debris removal has none, because the beneficiary of a removal is everyone and the payer is no one.

Frontier Screening at scale. A billion pairwise comparisons per run and a million predicted weekly conjunctions on bad days already outstrip human-in-the-loop operations; automation of risk assessment and manoeuvre negotiation (ESA’s CREAM line of work, operator autonomy like Starlink’s) exists but has no shared protocol, so each operator automates into a void.

Established Thermosphere density forecasting. Drag is both the great cleanser and the great error source: solar-cycle-driven density swings change reentry timelines by years and corrupt conjunction predictions within days. Progress rides on space-weather modelling of the kind surveyed in space weather engineering; the current solar maximum has measurably thinned debris at constellation altitudes while also degrading prediction accuracy.

Frontier Autonomous rendezvous and proximity operations. Removal, inspection and servicing share one technology base — relative navigation, safe trajectories around uncooperative targets, capture mechanisms — which also underpins orbital shipyards and space-based manufacturing. ADRAS-J is currently the best public evidence of its maturity.

Frontier Machine learning for screening and manoeuvre planning. Conjunction triage, covariance correction and multi-party manoeuvre optimisation are active research areas; the constraint is less algorithmic than data-access: the best training data sits in classified or proprietary silos.

Established Standards bodies. The CCSDS conjunction data message format, ISO 24113 (space debris mitigation requirements) and IADC guidelines are the load-bearing technical standards; every coordination proposal builds on them. Frontier Economics and institutional design — commons pricing, licensing harmonisation, liability reform — are dependencies in the strict sense: several bottlenecks above are unbuildable without them, a theme shared with space law and governance.

Frontier The decisive demonstration is JAXA’s CRD2 Phase II: Astroscale’s ADRAS-J2 is contracted to grapple the same H-2A upper stage with a robotic arm and drag it into a destructive reentry — a capture of a large, uncooperative, slowly tumbling object that nobody has ever performed. With a flight window announced for no earlier than 2027, it is the nearest-term test of the claim on which the whole remediation agenda rests: that removal is an engineering problem rather than a permanent cost sink. Success at a published price would let policy argue about who pays; failure or major overrun pushes remediation back a decade.

Frontier ClearSpace-1 is the independent replication. A different consortium, a different capture architecture (enclosing capture arms rather than a single manipulator), a different target (the PROBA-1 satellite, after the original VESPA adapter target was struck by debris), flying around 2028. Two successes by different teams would establish removal as a capability; two failures would be equally informative.

Established The running measurement is SpaceX’s semi-annual FCC filing series. It is the only longitudinal public dataset on traffic-management load: manoeuvre counts, warning-pipeline outages, disposal statistics. The near-term question it will answer: whether manoeuvres per satellite-year keep climbing as solar activity declines and Chinese constellations scale — Lewis’s projection of 80,000 per half-year by 2027 is a stated, checkable forecast.

Frontier TraCSS is the natural policy experiment, already running. The US is testing, in real time, whether a civil space-traffic authority can survive its own budget process: terminated in the FY2026 request, restored by Congress at $52.5 million, cut to $11 million in the FY2027 request, serving 62 organisations while formally still a pilot. Its resolution — institutionalised, privatised, or archived — will set the template other jurisdictions copy.

Established The model-falsification ledger is kept annually. ESA’s Space Environment Report records each year’s fragmentation events, launch traffic and disposal compliance; a decade of that series against DELTA- and LEGEND-class projections is the cleanest available test of whether the environment is tracking the stable, slow-growth or accelerating branch of the models.

Speculative The unrun experiment is economic. No jurisdiction has piloted an orbital-use fee, a disposal bond, or a remediation levy, so every claim about incentive-based solutions rests on theory. A single mid-sized licensing state adopting a bonded-disposal requirement would generate the first empirical data on flag-shopping and compliance cost — nobody has scheduled one.

Frontier Removal vehicles are hard servicing missions with worse targets. A capture craft needs relative navigation against an object that may be tumbling and sun-glinting, capture hardware tolerant of contact dynamics with a three-tonne body, autonomy that fails safe within metres of the target, and enough propellant to drag the combined stack to a controlled destructive reentry. ADRAS-J’s safe-abort event showed the autonomy stack being exercised for real.

Established Design-for-capture is cheap now and impossible later. Standardised grapple fixtures and docking plates on new satellites — Astroscale sells one, and some constellations fit them — convert every future removal from a research problem into an operation. Retrofitting the existing derelict population is not an option; that stock must be handled the hard way or left up for decades (vendor interest noted).

Established Traffic engineering means machine-to-machine coordination. The required stack is known: shared ephemeris exchange with covariance, agreed screening thresholds, automated conjunction triage, and a deconfliction protocol that returns a binding answer to “who moves” in minutes. Pieces exist — CCSDS message formats, Space-Track APIs, ESA’s CREAM automation work, SpaceX’s internal autonomy — but no end-to-end system, and TraCSS is the only funded civil attempt at one.

Frontier Propellant budgeting is becoming a design driver. At roughly 30–35 manoeuvres per satellite per year at current thresholds, avoidance consumes station-keeping margin; constellations must budget for it across five-year vehicle lives, and a satellite that cannot manoeuvre (100 such outages in one six-month Starlink report) becomes everyone else’s problem.

Frontier Disposal engineering has a fork. Demisable designs that burn up completely trade against controlled-reentry designs that survive to a targeted ocean impact; the first pollutes the upper atmosphere at scale, the second risks casualties if uncontrolled. Constellation-era disposal rates — 73 reentries in six months for one operator — make this a live design decision, not a paperwork item.

Established The legal scaffolding lives one brief over. Space law and governance owns the treaty layer this brief keeps colliding with: the Outer Space Treaty’s state-responsibility articles, the 1972 Liability Convention, and the registration regime that determines who may even touch a given derelict. Debris is that framework’s hardest stress test, because fault in orbit is unprovable under rules written for 1972.

Frontier Everything upstream depends on this working. Deep space infrastructure, orbital shipyards, space habitats and space-based solar power all assume cheap, routine transit through and operation in LEO; large structures are large cross-sections, and a shipyard is a conjunction magnet. Their economics quietly assume the traffic problem stays solved.

Established Sensor and RPO synergies run both directions. The survey networks that feed conjunction screening share hardware and catalogues with planetary defense; rendezvous-and-capture technology matured on debris targets transfers directly to satellite servicing and space-based manufacturing. Frontier Launch-side coordination — licensing cadence, collision-avoidance screens for ascent, reentry corridors — couples this topic to spaceports, and drag forecasting couples it to space weather engineering.

Established The core institutional fact: warning is a public good currently provided by a military organisation that wants to hand it off, and the hand-off keeps failing. The US Space Force screens the world’s traffic; SPD-3 (2018) ordered the civil transfer; TraCSS was beta by September 2024, then proposed for termination in June 2025, saved by Congress at $52.5 million, and cut to $11 million in the FY2027 request. Both the Space Force and an industry coalition of 450+ companies publicly opposed termination — an unusual alignment that still did not secure stable funding. The institutional form — agency, contractor-owned, or data-as-a-service with fees — remains explicitly undecided as of mid-2026.

Established The international layer is soft law plus one hard lever. IADC and UN COPUOS guidelines bind no one; ISO 24113 binds only those who contract to it; the EU’s Space Surveillance and Tracking consortium provides an independent European warning service. The hard lever is market access: the FCC’s 2022 five-year rule reaches any constellation that wants to sell in the US, making a national regulator the de facto global standard-setter — efficient, and resented.

Established Liability exists on paper and has settled exactly one claim. The 1972 Liability Convention makes launching states absolutely liable for surface damage and fault-liable in orbit. The only invoked case is Cosmos 954: Canada billed the USSR C$6 million for the 1978 scattering of a nuclear reactor across the Northwest Territories and settled for C$3 million in 1981. For an in-orbit collision, fault requires proving negligence against no agreed standard of care — no right-of-way rules, no binding manoeuvre obligations — so the regime deters nothing. Frontier Insurance mirrors this: third-party in-orbit collision cover is thin, priced on ambiguity, and unavailable at constellation scale.

Frontier Consent is a binding constraint on remediation. Under the treaty regime, jurisdiction over an object never lapses: removing another state’s derelict without agreement is unlawful, so the statistically optimal target list (heavy with Soviet-era stages) is mostly off-limits to Western missions. Every current removal mission targets the sponsor’s own national debris for exactly this reason.

Frontier Who pays is unresolved at every level. Within governments (Commerce vs Defense budgets), between operators (fewer than ten reciprocal coordination manoeuvres in a Starlink reporting period), and globally (no fee, bond, levy or fund exists anywhere). ESA’s Zero Debris Charter, with well over a hundred signatories, is the current best attempt to build a coalition of the willing ahead of binding rules.

Frontier Private rulemaking by first movers. When one operator flies more than half the active satellites, its internal choices — manoeuvre thresholds, autonomy behaviour, disclosure practice — become the de facto traffic code without any public process. SpaceX’s transparency is currently exemplary and entirely voluntary; the ethics problem is the “voluntary,” not the behaviour.

Established Latecomer equity. Orbital shells are appropriated by occupation: filings totalling tens of thousands of satellites from a handful of states and companies are claiming the best LEO real estate while the guidelines remain non-binding. States that will be able to afford constellations in 2050 will inherit whatever risk environment 2020s operators leave, with no compensation mechanism — the intergenerational structure of every commons problem.

Established Ground casualty and atmospheric costs are real but small versus attention paid. Reentry casualty expectation per event is regulated to 1-in-10,000; with disposal rates now in the hundreds per year, aggregate exposure and stratospheric metal deposition both rise, and the 2024–2025 atmospheric-chemistry literature treats the latter as a genuinely open question. Frontier Astronomy bears a separate uncompensated cost in streaked images and radio interference.

Established Debris is also a weapons-norms problem. The worst single debris events are anti-satellite tests. The US declared a destructive direct-ascent ASAT test moratorium in 2022 and a UN General Assembly resolution endorsed it; China and Russia have not joined. Any traffic-management regime is hostage to one afternoon of testing.

Established LEO is critical infrastructure now, not eventually. Navigation, timing, weather, communications, missile warning and increasingly consumer broadband run through orbits whose management still depends on email, goodwill and one military squadron’s screening runs. The civilizational exposure is not losing space; it is the quiet accumulation of cost and fragility in systems everything else sits on.

Frontier The realistic bad case is a degraded commons, not a sealed planet. The models’ bad branch looks like: certain shells become uninsurable, manoeuvre budgets and shielding grow, small operators are priced out, and launch windows and disposal rules tighten — a tax on every orbital activity, compounding for decades. That outcome forecloses nothing absolutely but makes every ambition on this map — from asteroid mining to lunar industry, all staged through LEO — permanently more expensive.

Speculative The governance precedent may matter more than the orbits. Orbital traffic is the first genuinely global, fast-moving commons humanity is trying to manage with real-time data and mixed public-private authority. The institutions built (or not built) here — fees, right-of-way, remediation obligations — are the likely template for later commons: spectrum, cislunar space, perhaps geoengineering. Getting it wrong teaches the wrong lesson at civilizational scale.

These horizons track the coordination and remediation machinery, taking the measured manoeuvre and population record as the baseline.

• 10 yr: Frontier The first robotic captures of large debris succeed or fail (ADRAS-J2, ClearSpace-1); the US settles TraCSS’s institutional form; manoeuvre counts reach the high hundreds of thousands per year as Chinese constellations scale; the FCC five-year rule produces its first enforcement wave; atmospheric effects of mass reentry get their first quantitative constraints.
• 25 yr: Speculative Removal becomes a priced service bought by agencies and insurers at perhaps a few large objects per year — if capture demonstrations succeed and a payer is constructed; machine-to-machine deconfliction with agreed right-of-way conventions covers the major constellations; some shells operate under explicit capacity management.
• 50 yr: Speculative Debris-neutral operation (Zero Debris-style) is the licensing norm in major jurisdictions; the legacy derelict stock is being drawn down at the five-to-ten-per-year rate the models ask for, or has been accepted as a permanent background tax; the environment’s trajectory is unambiguous in the annual measurement series either way.
• 100 / 250+ yr: Handwave Orbital zoning with enforced lanes, in-orbit recycling of derelict mass as feedstock for orbital shipyards, and self-funding remediation utilities — each assumes institutional inventions nobody has designed and an orbital economy large enough to want them.

• Depends on little on this map: measuring the environment and coordinating traffic are blocked by institutions and money, not by missing science elsewhere in the corpus. The useful upstream inputs are drag and solar-activity forecasting from space weather engineering and the treaty-reform agenda in space law and governance.
• Requires (not on this map) a measured, not modelled, flux for the lethal 1–10 cm population; capture vehicles that can reliably grapple uncooperative tumbling rocket bodies; a binding right-of-way and deconfliction protocol so that “who moves” has an answerable form; a constructed payer — fee, bond, fund or mandate — for removal services; and working conjunction-coordination channels with Chinese constellation operators, currently an email address.
• Enables every LEO-transiting ambition at stable cost: deep space infrastructure, orbital shipyards, space habitats, space-based solar power. It is an enabling layer in the strict sense: its success is invisible and its failure is priced into everyone else’s missions.
• Adjacent to space law and governance (the liability and consent regime), planetary defense (shared survey infrastructure), and spaceports (launch and reentry corridor coordination).

Handwave “The Kessler syndrome will suddenly cascade and trap us on Earth.” The popular version — hours of chain-reaction destruction, as in the film Gravity — is not what the models or Kessler himself say. Kessler’s 1978 paper described a slow statistical feedback; he stated explicitly that cascading “was never intended to mean” a process of days or months, and in 2012 described it as “continuous and as already started,” each collision slowly raising the frequency of the next. The realistic failure mode is decades of compounding cost, not a wall.

Handwave “Space is big; collisions are a negligible risk.” The manoeuvre record refutes this from the other direction: an operator flying at conservative thresholds executed roughly 300,000 avoidance manoeuvres in one year. Orbits are not volume; they are intersecting lanes in a handful of preferred shells, and the 2009 Iridium–Cosmos collision happened at the intersection of exactly two such lanes.

Frontier “Megaconstellations are the debris problem.” Partly backwards. Tracked fragment counts and on-orbit mass are dominated by legacy rocket bodies, dead satellites and ASAT tests; Starlink satellites fly low, deorbit on failure within years, and dispose at unprecedented rates. What constellations dominate is the traffic and coordination load — and they concentrate collision consequence in specific shells. Both clauses are true; conflating them misdirects policy.

Established “Tracked means we know where things are.” A catalogue entry is an orbit with uncertainty, commonly hundreds of metres to kilometres; a single manoeuvre can shift predictions by tens of kilometres for days. Conjunction warnings are probability statements over fat error ellipsoids, which is why thresholds, not detections, drive manoeuvre counts.

Frontier “Removal technology exists; only funding is missing.” No large uncooperative object has ever been removed. Inspection has flown (ADRAS-J); capture of prepared targets has flown (ELSA-d); capture of a real tumbling derelict has not, and the first two attempts are still ahead of their launch dates. Funding is missing too — but so is the demonstration.

Established “The 25-year rule is law.” It is a guideline, historically honoured unevenly; the binding instruments are national licences. The FCC’s 2022 five-year rule binds only satellites touching the US market, and the first-ever debris fine (2023) was $150,000 — enforcement exists, at the scale of a rounding error.

Established “Liability law will discipline behaviour.” One claim in five decades, settled diplomatically (Cosmos 954), and an in-orbit fault standard nobody can meet because no rules of the road exist against which to be negligent. Operators behave well, where they do, for self-interest and reputation, not legal exposure.

Speculative “Lasers will sweep the sky clean.” Ground-based photon-pressure nudging of small debris is a real research line and a plausible eventual tool for the 1–10 cm class that capture missions cannot touch; no system has moved a single piece of orbital debris to date, and power, pointing and liability questions (a laser that can move debris can blind satellites) are all open.