Megaconstellation Externalities

Established A satellite constellation is a machine for turning private capital expenditure into physical changes in shared media that nobody bills for. Five of those media are now separately measurable: the optical night sky, the radio spectrum in bands reserved for passive science, the mesosphere and stratosphere into which dead satellites are deliberately vaporised, the occupancy of particular altitude shells, and the debris footprint of uncontrolled reentry. This brief is about what happens when you add them together.

Established The scale changed over six years, not sixty. The active satellite population went from roughly 2,000 spacecraft at the end of 2019 to more than 10,000 by early 2026, over half of them belonging to a single operator, and orbital launch attempts set records in 2024 and again in 2025. Filings lodged with the International Telecommunication Union across some 300 proposed constellations sum to well over one million satellites — a number that reflects the near-zero cost of filing rather than any deployment plan, but that defines the legal envelope inside which operators claim priority.

Established This brief deliberately does not re-argue collisions, tracking or debris removal. Conjunctions, the debris population, active removal and the fragility of civil traffic services belong to Orbital Debris and Space Traffic, which lands those numbers: the tracked catalogue, the 300,000 avoidance manoeuvres logged by one operator in 2025, the removal demonstrations that have never grappled a large derelict. Here the traffic layer is given, and the question is what a constellation does to everything that is not a collision ledger.

Frontier The synthesis claim is about arithmetic, not outrage. Every externality here is small enough for a licensing authority to dismiss on its own: a few per cent of exposures spoiled, excess noise in one band, a tonne a day of metal in a stratosphere that already receives meteoric metal. Summed across media and across the announced fleet, they are a measurable, uncompensated transfer from scientific and cultural uses with no property rights to commercial uses with licences. No jurisdiction requires an applicant to compute that sum. The gap between what is measurable and what is countable is the subject.

Established A note on sourcing. This brief was commissioned in September 2026 from the Institute’s research base. Reading-list entries without links are cited from the bibliographic record rather than re-fetched, and claims are dated no later than early 2026 unless carried by a linked source.

Established Satellite brightness is a photometric quantity and it has been measured repeatedly by independent observers. First-generation Starlink spacecraft at their 550 km operational altitude were measured at a mean apparent magnitude near 5.5, naked-eye visible under dark skies and far brighter than the level at which sensitive detectors saturate. A darkened coating flown on one spacecraft cut brightness by roughly 1.5 magnitudes but failed on thermal control; a deployable sunshade on the next generation delivered about 1.3 magnitudes and went into series production. The mitigations were the operator’s choices, but the photometry that graded them came from independent observers, which is why the record is credible.

Established The astronomical community set a numerical target and it is mostly not being met. Workshops convened by the International Astronomical Union and the US national observatory system converged on apparent magnitude 7 at a 550 km reference altitude as the level below which most research programmes can tolerate a satellite crossing the field. Second-generation spacecraft are physically larger — heavier buses, bigger arrays, in some designs a direct-to-cellular panel of tens of square metres — and unmitigated versions measure brighter than the first generation despite the earlier fixes. Dielectric mirror films and array off-pointing have pulled some units toward the target; the population median has not reached it.

Established The worst measured case is not a broadband constellation at all. A prototype direct-to-handset spacecraft carrying a phased-array panel of roughly 64 square metres was measured by a coordinated international campaign at a peak apparent magnitude of about 0.4, among the brightest objects in the night sky when overhead. That measurement, published in Nature in 2023, is the most consequential brightness result in the field: the driver is deployed area rather than satellite count, and the direct-to-handset business case requires large area by construction.

Frontier The cost to astronomy is quantified for narrow-field instruments and contested for wide-field surveys. Modelling of the European Southern Observatory’s facilities found that for narrow-field, long-exposure instruments the fraction of exposures ruined is small — low single-digit percentages, concentrated in twilight — and largely mitigable by scheduling. For wide-field imagers the picture inverts: pre-survey predictions for the Vera C. Rubin Observatory put the fraction of twilight-band images carrying at least one trail in the tens of per cent under a fleet of tens of thousands, with residual detector artefacts — crosstalk, ghosts, saturation trails — persisting after the streak is masked. Those predictions were published before the survey began, which is what makes the coming years informative.

Frontier Diffuse skyglow from the aggregate population is a real prediction and an unmeasured quantity. Unresolved sunlight scattered from the whole orbiting population — satellites plus debris too small to track — raises diffuse night-sky brightness everywhere, including over sites protected from ground-based light pollution. A widely cited 2021 radiative-transfer estimate put that increase near 10 per cent above the natural background, which is the threshold the IAU has used since 1979 to define an unpolluted site. The physics is uncontroversial; the number depends on assumed albedo, size distribution and scattering phase function for a population nobody has characterised optically, and separating a 10 per cent diffuse term from airglow variability is genuinely hard.

Established Radio interference from constellations is measured, and the dominant channel is not the downlink. Observations with the Low Frequency Array in the Netherlands detected broadband unintended electromagnetic radiation from a large fraction of the Starlink spacecraft observed, spanning 110–188 MHz including the band 150.05–153 MHz that the Radio Regulations reserve for radio astronomy. This emission does not come from the communications payload operating in its allocated band; it leaks from onboard electronics — power converters, avionics, propulsion controllers. A follow-up campaign found second-generation spacecraft radiating substantially more strongly than the first generation across the same range, with flux densities orders of magnitude above the sensitivity floor of next-generation low-frequency arrays.

Established There is no binding rule against unintended emission from a spacecraft. The ITU Radio Regulations constrain what a licensed transmitter emits in and adjacent to its band, and ITU-R recommendations set thresholds of detrimental interference for radio astronomy. Spurious broadband radiation from a spacecraft’s internal electronics falls outside both: it is not an emission of a licensed transmission, an orbiting satellite is not a terrestrial device subject to national electromagnetic-compatibility law, and the protection criteria are recommendations. The practical consequence is that the most sensitive low-frequency array ever built sits inside a legislated terrestrial radio-quiet zone that offers no protection against an emitter passing overhead.

Established The atmospheric externality has the strongest measurement in the whole subject. In-situ sampling from high-altitude aircraft found that roughly one particle in ten of the sulfuric-acid aerosol in the lower stratosphere carried aluminium, lithium, copper, niobium and other metals in elemental ratios matching spacecraft alloys rather than meteoric infall. That result, published in the Proceedings of the National Academy of Sciences in 2023, is an instrument reading rather than a projection, and the same paper projects the metal-bearing fraction rising toward half of all stratospheric aerosol particles as the reentry mass flux grows with the fleet.

Frontier The mass flux is large and rising, and the ozone consequence is modelled, not observed. A constellation with a five-year design life and thousands of members must deorbit one to two thousand spacecraft a year in steady state; at several hundred kilograms each that is of order one to two tonnes a day of metal vaporised in the mesosphere. Modelling puts alumina injection near 17 tonnes a year at the 2022 reentry rate, rising several hundred per cent above natural meteoric alumina under full deployment of announced constellations. Alumina is a catalytic surface for chlorine activation — the mechanism that made polar ozone loss possible — and has stratospheric residence times measured in years. But no measured ozone anomaly has been attributed to reentry, the chlorine reservoir is declining under the Montreal Protocol, and the modelled effect sits inside the noise of current trend detection. The honest statement is that a plausible, irreversible, globally distributed perturbation is being run without a monitoring plan.

Established Licensing is national, additive and blind to the aggregate. The US Federal Communications Commission granted a second-generation constellation authorisation for 7,500 spacecraft against an application for 29,988; a competing operator holds authorisation for 3,236; Chinese national and provincial constellations have filed for tens of thousands more. Each grant is assessed against that applicant’s own orbital-debris showing, never against a cumulative optical, radio or atmospheric budget, because no such budget exists in any rule. A US Government Accountability Office review found the licensing agency applying a categorical exclusion from environmental review, dating from 1986, to commercial satellite authorisations, and recommended it revisit that position; the exclusion has been litigated rather than replaced.

Frontier The benefit side is real and should be stated at the same precision. Low-Earth-orbit broadband reaches users terrestrial networks were never going to reach economically: rural and remote households, maritime and aviation connectivity, disaster response, and wartime communications resilience. The connectivity case is documented in Next Generation Networks. Any honest accounting sets the measured costs against that, and the measured costs are not obviously larger. What they are is unpriced, unmonitored, and irreversible on different timescales than the benefits.

Frontier Does brightness mitigation scale with satellite area? Every demonstrated mitigation — darkening, sunshades, dielectric films, array off-pointing — cuts reflected flux by a roughly fixed factor, while deployed area grows. If that holds, the mitigated brightness of a 2030 spacecraft may exceed the unmitigated brightness of a 2020 one. No published design study shows magnitude 7 achievable for a 60 square-metre spacecraft.

Frontier What is the atmospheric fate of spacecraft-derived alumina and lithium? Particle size distribution, altitude of formation, coagulation with sulfate, residence time and the effective surface area available for heterogeneous chemistry are all model assumptions calibrated on solid-rocket-motor plume studies from a different altitude regime. The measured composition record exists; the microphysics does not.

Speculative Does deliberate reflection become a distinct category? Proposals to fly orbital mirrors directing sunlight at ground targets invert the mitigation problem: brightness is the product, not a defect. The lineage runs back to Soviet mirror experiments in the 1990s. Whether any such system reaches operational scale is unresolved; the licensing question is whether an intentionally bright object is treated differently from an incidentally bright one, and no regime distinguishes them.

Frontier Who has standing to complain? Astronomers are not licensees, hold no spectrum, and in most jurisdictions cannot show the particularised injury administrative review requires. It is a legal-doctrine question and the largest single determinant of whether any measured effect ever gets priced.

Established There is no continuous, calibrated, global brightness monitoring of the satellite population. Photometry comes from a patchwork of amateur observers, occasional professional campaigns and operator self-reporting. A regulator wanting to enforce a magnitude limit would have no instrument of record. Building one is neither expensive nor hard — a few dozen automated wide-field cameras with photometric standards — and nobody has funded it.

Established Unintended radio emission can only be measured by the instruments it damages. The detections came from research arrays observing in their own science bands. There is no dedicated monitoring facility, no pre-launch test standard for on-orbit unintended radiation, and no requirement that an operator characterise its bus before flight. Terrestrial electromagnetic-compatibility testing is routine; extending it to an orbital emission budget is an administrative decision, not a technical one.

Frontier Stratospheric composition sampling is aircraft-limited. The metal-particle result depends on a small number of high-altitude research aircraft flown on campaign schedules. A time series able to detect a rising trend against seasonal and volcanic variability needs a committed cadence, which no agency has scheduled; balloon and satellite-limb alternatives have lower composition specificity.

Speculative Demisable materials that do not produce catalytic aerosol are an unexplored design space. The requirement to fully demise and the requirement not to inject alumina point at different alloy choices. Whether a structural material exists that satisfies both, at satellite cost and mass, has not been seriously studied because nobody has posed the requirement.

Established Everything optical depends on accurate ephemerides, which come from the traffic layer. Trail avoidance by scheduling and post-hoc masking both require knowing where every satellite will be to arcsecond precision minutes ahead. That capability is a by-product of the tracking infrastructure described in Orbital Debris and Space Traffic, and inherits its funding fragility: a mitigation strategy built on operator-supplied ephemerides is only as durable as the arrangement supplying them.

Established Atmospheric effects depend on solar-cycle forecasting. Thermospheric density drives both derelict lifetime and reentry rate, so the timing and altitude of metal injection are functions of solar activity. The forecasting problem and its error bars belong to Space Weather Engineering; the February 2022 storm that destroyed dozens of newly launched satellites during orbit raising shows how large the coupling is.

Frontier Ozone attribution depends on the trend-detection literature. Claiming a reentry-driven ozone effect requires the statistical machinery for detecting small trends in a noisy, recovering system — the same machinery the Atmospheric Management brief shows to be fragile, where instrument-network stability and sonde calibration limit what can be resolved. A reentry signal is expected to be smaller than the residuals in that record.

Frontier The decisive test is already running: the Rubin Observatory’s survey statistics against its own pre-survey predictions. Predictions for the fraction of images contaminated by trails, and for residual damage after masking, were published before the ten-year survey began, at stated assumptions about fleet size and brightness. The survey now produces that measurement directly, in one consistent pipeline, at a cadence no other facility matches. A contamination rate near the low end would show scheduling and masking work and would deflate the astronomy externality; a rate near the high end, or residual artefacts that masking does not remove, would convert a modelled loss into a documented one and give the astronomical community the single number it has never had.

Frontier The strongest unscheduled measurement is a stratospheric metal-fraction time series. The in-situ aerosol sampling that found spacecraft metals in one particle in ten produced, in the same paper, a falsifiable projection of that fraction rising toward one in two. Re-flying the same instrument on the same aircraft at a fixed cadence for a decade would test it directly, needs no new hardware, and would be the first empirical trend in any megaconstellation externality. Nobody has committed to the cadence.

Frontier The radio experiment is a pre-flight emission characterisation. No operator publishes a broadband unintended-emission budget for its bus. One manufacturer measuring and publishing one, and one low-frequency array re-observing the same spacecraft class before and after a shielding change, would establish whether this is a design defect costing grams to fix or an intrinsic property of high-power spacecraft electronics. The technique is proven; a cooperating operator is not.

Frontier The natural policy experiment is the direct-to-handset generation. Several operators are deploying spacecraft whose business model requires very large area in low orbit. Their measured brightness against the magnitude-7 guideline, and whether any licensing authority conditions an authorisation on it, tests whether voluntary astronomy coordination has force. The first authorisations and photometry are arriving now, a running experiment with a short reporting time.

Established Optical mitigation is a solved engineering problem at small scale and an unsolved one at large scale. The toolkit is known and each element has flown: low-reflectance coatings, deployable visors, dielectric films that redirect specular reflection away from the ground observer, array orientation control during orbit raising, and lower operating altitudes. What has not been demonstrated is any combination delivering magnitude 7 on a spacecraft with tens of square metres of deployed area, and the thermal budget binds: a surface that does not reflect sunlight absorbs it.

Frontier Radio mitigation is conventional electromagnetic-compatibility engineering applied to an unconventional requirement. Shielding, filtering, layout discipline and switching-frequency selection are standard practice on the ground because terrestrial equipment must pass a compliance test; applying them to a satellite bus costs mass and money and nothing else. The open question is the target level, since a protection criterion expressed as spectral flux density at a ground array must be back-propagated into a component-level emission budget through an antenna pattern nobody has characterised.

Frontier Atmospheric mitigation runs against the disposal requirement. Controlled reentry into an ocean disposal zone concentrates mass at low altitude rather than aerosolising it, at the price of retained propellant and therefore payload. Graveyard disposal defers the problem and conflicts with post-mission disposal rules. Material substitution — structures that demise into species with no known catalytic chemistry — is the interesting option and the unstudied one. Recovery and reuse in orbit, discussed in In-Space Servicing and Depots, would eliminate the flux entirely and is nowhere near the cost point.

Established The traffic and debris layer is the sibling brief, not a subsection of this one. Orbital Debris and Space Traffic owns conjunction screening, the debris population, active removal and the institutional survival of civil traffic services. The two briefs share one causal link worth naming: fragmentation events increase the small-debris population, and small debris contributes to the same diffuse skyglow term as the satellites themselves, so a collision is simultaneously a traffic event and an optical one.

Established Observational astronomy is the injured party and its economics are covered elsewhere. Mega-Telescopes sets out the cost scaling and funding fragility of the facilities at stake; Multi-Messenger Astronomy describes the time-domain follow-up chains that depend on wide-field survey alerts, the observing mode most exposed to trails; and Planetary Defense relies on the same twilight survey capability, which puts a public-safety function on the cost side of the ledger.

Established The atmospheric chemistry connects to two existing briefs. Atmospheric Management supplies the ozone trend-detection and sink-chemistry machinery that any attribution claim must pass through, and Artificial Satellites for Climate Control covers the deliberate-reflection proposals that would make orbital brightness a design objective rather than a defect.

Established The regulatory frame sits in the space-law brief. Space Law and Governance establishes what the Outer Space Treaty’s consultation provisions actually oblige, which is very little, and why national licensing rather than international law is where any constraint will originate. Spaceports and Next Generation Networks carry the launch-cadence and connectivity-benefit sides of the same system.

Established Three bodies matter and none of them has the mandate. The International Telecommunication Union allocates spectrum and coordinates interference between radio services; it has no competence over optical brightness or atmospheric chemistry, and protects radio astronomy through recommendations rather than binding obligations for the emission class that causes the problem. National licensing authorities grant market access and impose debris conditions, one applicant at a time. The UN Committee on the Peaceful Uses of Outer Space has taken up dark and quiet skies as an agenda item; it works by consensus and issues guidelines.

Established The astronomical community built the institution the regulators did not. The IAU established a Centre for the Protection of the Dark and Quiet Sky from Satellite Constellation Interference in 2022, co-hosted by the US national optical observatory and the international radio-astronomy organisation. It produced the brightness recommendations, ran the photometric campaigns behind the key measurements, and negotiated directly with operators. Its entire authority is reputational. That it has achieved anything is evidence voluntary coordination is not worthless; that the population median still misses the target is evidence about its ceiling.

Frontier Environmental review is the live legal front. The US licensing agency’s use of a categorical exclusion to avoid environmental analysis of satellite authorisations, criticised by the Government Accountability Office, is the one hook through which cumulative effects could enter a licensing decision, because environmental statutes generally require consideration of cumulative impacts where individual ones are small. Whether orbital and upper-atmospheric effects count as the environment for statutory purposes is unresolved and jurisdiction-specific. A decision either way would be the most consequential institutional event in this subject.

Speculative The instrument that does not exist is a cumulative sky-burden test. The design is not mysterious: an applicant computes its contribution to defined indices — integrated reflected flux in a standard band, spectral flux density in protected radio bands, annual mass of demisable material by species — and the authority compares the running total to a stated ceiling. Every element is measurable today. No jurisdiction has proposed it, and the first to do so would meet an immediate competitiveness argument from its own launch industry.

Established The night sky is a cultural and not only a scientific resource, and the loss is unevenly distributed. Naked-eye satellite visibility is worst at latitudes near 50 degrees and in the hours after dusk and before dawn, because that is where and when the shells stay sunlit. Astronomical practice embedded in Indigenous and traditional knowledge systems depends on specific horizon and twilight observations, which are the observations most affected, and those communities are largely not the ones deciding about licences.

Frontier The benefit distribution runs the other way, which is why the ethics is hard. Remote connectivity, maritime and aviation safety communications, and emergency service after infrastructure failure accrue disproportionately to under-served populations. A brief treating this as commercial vandalism would be wrong. The defensible position is narrower: the benefit is priced and the cost is not, so nobody ever actually makes the tradeoff, and a system that never makes it will over-supply the activity wherever the true optimum lies.

Frontier Irreversibility differs by medium and that should drive precaution differently. Optical and radio effects stop when the satellites stop: a constellation abandoned tomorrow leaves a sky that recovers on the orbital-decay timescale of years to decades. Stratospheric metal loading persists on aerosol residence timescales and, if it perturbs ozone chemistry, on the recovery timescale of the ozone layer — many decades. The atmospheric externality therefore deserves a stricter standard of proof-of-safety than the optical one, and currently receives a looser one.

Frontier This is the cleanest available test case for governing a new commons before it is spoiled. The externalities are physical, measurable with existing instruments, attributable to identifiable actors, and small enough today that mitigation is cheap. If a civilisation cannot price an externality with those properties, the prospects for harder ones — lunar resources, cislunar traffic, deep-space contamination — are worse. This is the easy version of the problem, and it is being failed.

Speculative The observational consequence compounds over centuries if the trend continues. Ground-based optical astronomy has a long-run answer — the lunar far side, or orbits above the constellation shells — at one to three orders of magnitude more than the ground facility it replaces. If the externality is never priced, the outcome is a slow transfer of astronomy off Earth at public expense: a subsidy from science budgets to satellite operators, paid through a channel nobody set up on purpose.

Handwave The claim that constellations foreclose humanity’s view of the universe is not supported. Radio astronomy at protected frequencies, infrared and submillimetre astronomy, space-based observatories, gravitational-wave and neutrino observatories are all essentially unaffected by satellite reflection, and the affected optical modes are affected fractionally rather than absolutely. The real civilizational stake is narrower and more mundane: whether a specific class of wide-field, time-domain, twilight observation — including the one that finds Earth-crossing asteroids — survives at usable quality.

These horizons track when each externality becomes measurable, when it becomes priced, and when either is reversed.

• 10 yr: Frontier The Rubin survey returns a measured trail-contamination rate against its own predictions; direct-to-handset constellations are photometered at scale; at least one jurisdiction litigates whether satellite licensing requires environmental review. A binding brightness limit in any national licence remains unlikely; a voluntary emission-budget disclosure by at least one manufacturer is plausible. Stratospheric metal fraction is re-measured at least once, probably not as a committed series.
• 25 yr: Speculative Either a cumulative sky-burden test exists in one major licensing regime and is being copied, or the question has been settled by default and the astronomy community has redirected capital toward space-based and far-side facilities. Stratospheric metal loading is either a documented trend with a policy response or a documented trend without one; the physics makes the trend itself near-certain if the fleet grows as filed.
• 50 yr: Speculative The orbital environment supports either a governed population with per-shell occupancy limits and internalised costs, or a much larger ungoverned one whose externalities are absorbed as a background condition. The first path requires an institution nobody has built; the second requires no decisions at all, which is why it is the default.
• 100 / 250+ yr: Handwave Projections at this range extrapolate an industry six years old and are scenario sketches. If orbital manufacturing and servicing mature, reentry flux could fall to near zero because mass becomes too valuable to burn; if not, the mesosphere is a permanent industrial waste stream. Nothing in the evidence base discriminates between these.

• Depends on the tracking and ephemeris layer described in Orbital Debris and Space Traffic, without which neither trail avoidance nor post-hoc masking works; the drag and solar-cycle forecasting in Space Weather Engineering, which sets reentry timing and therefore the atmospheric flux; and the ozone trend-detection machinery assessed in Atmospheric Management, which bounds what any attribution claim can say.
• Requires (not on this map) an observed, attributed ozone response to reentry-derived alumina rather than a modelled one; a licensing test in at least one major jurisdiction that scores an applicant’s cumulative optical, radio and atmospheric contribution rather than its debris plan alone; a binding limit on unintended broadband emission from spacecraft electronics, which today falls outside every instrument; a price or fee attached to occupancy of a specific altitude shell, so that shell crowding has a cost to the occupier; satellite buses with tens of square metres of deployed area that still meet the magnitude-7 guideline, which no published design achieves; and structural materials that satisfy design-for-demise while producing no catalytically active aerosol, a design space nobody has surveyed.
• Enables a defensible answer to the question every larger orbital programme will eventually be asked — what does this do to the shared environment, and who pays for it. A working externality-accounting regime in low Earth orbit is the template that Space-Based Solar Power, Orbital Shipyards and Space Habitats would each be assessed under, and its absence is a licensing risk those programmes currently do not price.
• Adjacent to Mega-Telescopes and Multi-Messenger Astronomy on the injured side, Next Generation Networks on the benefit side, Space Law and Governance for the consent and liability regime, and Artificial Satellites for Climate Control where orbital reflection becomes deliberate.

Handwave “Megaconstellations are ending ground-based astronomy.” The numbers do not support it. For narrow-field large telescopes the loss is a low single-digit percentage of exposures concentrated in twilight and largely recoverable by scheduling; for wide-field surveys the contaminated fraction is much larger but masking recovers most affected pixels. What is genuinely at risk is one observing mode — wide-field twilight imaging for fast transients and inner-solar-system objects — and it deserves to be argued on its own terms rather than inflated into the death of a discipline.

Handwave “Satellite reentries are destroying the ozone layer.” No ozone anomaly has been attributed to reentry. The chain from measured stratospheric metal particles to ozone loss runs through several modelling steps — particle size and surface area, chlorine availability, injection altitude — each with real uncertainty, and the modelled effect sits below the detection threshold of the monitoring record. The defensible claim is precautionary: a persistent, irreversible, globally distributed chemical perturbation is being introduced without a monitoring programme, which is bad practice independent of the eventual answer.

Established “Burning up on reentry is clean disposal.” This confuses ground casualty risk with environmental effect. Design-for-demise requirements exist to prevent fragments reaching the ground and they work; their direct consequence is that a larger fraction of each spacecraft is converted to metal aerosol in the upper atmosphere. The two goals conflict and no regulation acknowledges the conflict.

Established “There are international dark-and-quiet-skies rules.” There are recommendations, workshop reports, a UN agenda item and a voluntary coordination centre. No binding instrument anywhere requires an operator to limit satellite brightness, and the radio protections that exist do not cover the unintended broadband emission the measurements actually found. Documents citing international guidelines here are citing guidance no one is obliged to follow.

Frontier “Atmospheric drag cleans up low orbits automatically, so nothing accumulates.” True for objects, and the main reason the low shells are not already a debris trap. It is also the mechanism that delivers the metal into the atmosphere: citing drag as the reason not to worry about orbital accumulation is citing the reason to worry about stratospheric accumulation.

Handwave “Market forces will sort it out, because operators need dark skies too.” They do not. No operator’s revenue depends on the optical or radio quality of the night sky, on stratospheric aerosol composition, or on a public observatory’s efficiency. This is the textbook structure in which private and social optimisation diverge, and asserting otherwise is where the case for doing nothing runs purely on assertion.