Mega-Telescopes

Mega-telescopes are the next generation of extremely large observatories on the ground and in space: 25-to-39-metre segmented mirrors in the Chilean Atacama, a 6.5-metre infrared observatory already working at the second Lagrange point, a wide-field survey telescope that launched in August 2026, and a proposed flagship designed to photograph an Earth-like planet beside its star. The framing under test on this page is that ever-larger telescopes follow from ever-better engineering.

Established This cluster refutes that framing more decisively than any other subject in the category, and it refutes it with something close to a controlled experiment. Two US giants of comparable design maturity entered the same decade — the same segmented-mirror technology, the same adaptive-optics lineage, international consortia on both, published cost estimates within about forty per cent of each other. One was advanced and one was dropped, and the discriminator was site access and a budget that could carry only one.

Established The single most useful number on this page is an exponent. The recommended planning relation for giant telescope construction cost is that cost scales as roughly the 2.7 power of diameter. That says doubling the aperture costs about six and a half times as much — not four times, as collecting area alone would suggest, and certainly not twice. Frontier On that exponent a 100-metre ground telescope would cost roughly fourteen times the 39-metre one, which is arithmetic on a published heuristic rather than a costing, and which is the whole reason the 100-metre concept of the early 2000s was abandoned in favour of a 39-metre.

Established One fact on this page changed two days before it was written and the tense had to change with it. The Nancy Grace Roman Space Telescope lifted off at 07:26 EDT on 30 August 2026 from Kennedy Space Center, on a three-month cruise to its final orbit with first images expected in early 2027. It flew. The coronagraph technology on which the starlight-suppression case for the next flagship rests is now an instrument in space rather than a plan.

Frontier And the flagship behind it has a budget line that should be read before anything else about it. The Habitable Worlds Observatory was ranked first by the last decadal survey, placed in a technology maturation programme in 2023, and had industry technology contracts awarded in January 2026. It was funded at $150 million in the enacted FY2026 budget and requested at $5 million for FY2027 — a single-year reduction of about 97%, which is arithmetic on those two figures.

Established What is actually under construction, starting with the largest. The European Southern Observatory's Extremely Large Telescope at Cerro Armazones in Chile has a 39.3-metre primary mirror of 798 hexagonal segments, each about 1.45 metres across and 50 millimetres thick. Its correction chain is six laser guide star units and a 2.4-metre deformable quaternary mirror with up to 5,000 actuators, readjusted a thousand times per second, for a claimed improvement of roughly 500 times in resolution over the uncorrected case. Cost: €1.055 billion for the downsized 2011 design and €1.15 billion including first-generation instruments as of 2017.

Established Its schedule is the finding rather than its size. The 2011 plan targeted first light in 2024–2025. The current schedule is technical first light in March 2029 and first scientific observations in December 2030 — roughly a five-year slip against the design-freeze plan, from a project that passed its halfway point in July 2023. Frontier Both schedules come from the same organisation, which is the interested party on its own project, and the second one is the reason to weight the first accordingly.

Established The Giant Magellan Telescope is the other architecture entirely. Seven 8.417-metre primary segments — one central and six symmetric — give a 25.4-metre aperture and 368 square metres of collecting area, with expected resolving power around ten times Hubble's and four times that of the current infrared flagship. All seven primaries had been cast as of 2023, with segments one to three complete and four to six in polishing, the first adaptive secondary under construction and the mount in manufacture. Cost estimate about $2 billion; site preparation began 23 March 2012, formal construction in November 2015, and commissioning is anticipated in the early 2030s. Frontier Those figures are the consortium's own, on its own project.

Established The Thirty Meter Telescope is the third, and its construction is halted. A 30-metre aperture of 492 segments, planned for Mauna Kea, with a design as complete as either of the others and no site it can build on.

Established The cost-scaling law is the answer to “why not just build it bigger,” and it is a published planning heuristic rather than a physical law. Cost scales as approximately the 2.7 power of diameter, attributed to work by Stepp and colleagues at the US national optical observatory. Fetched cost estimates sit reasonably on that curve: a large binocular telescope at about $120 million, the Giant Magellan at about $1,000 million, the Thirty Meter at about $1,400 million, and the Extremely Large at about $1,590 million. Frontier Those figures come from different years and currencies and were not normalised in the source, so they illustrate the exponent and should not be used to rank projects.

Established The completed space case is the one whose cost history carries the lesson. The late-1990s concept for what became the current infrared flagship carried a $1 billion price tag over ten years of construction, targeting a 2007 launch. In July 2011 a House appropriations committee moved to cancel it, describing it as billions of dollars over budget and plagued by poor management; Congress reversed the decision and capped additional funding to completion at $8 billion. The final figures: about $9.7 billion in NASA lifetime cost, of which $8.8 billion on design and development and $861 million for five years of operations, plus about €700 million from ESA and CA$200 million from Canada. It launched 25 December 2021 and reached its operating point on 24 January 2022, with a 6.5-metre primary of 18 gold-coated beryllium segments and 25.4 square metres of collecting area.

Frontier Read that history correctly and this page becomes much more useful than a cautionary tale. Roughly a factor of ten in cost and fourteen years of slip, both derived from the fetched figures — and the telescope works, spectacularly. The relevant lesson for the next flagship is not that space telescopes overrun. It is that a project of this class requires a political system willing to absorb a tenfold overrun and a near-death cancellation vote, across four administrations. That is the funding constraint stated as history rather than as worry.

Established Now the consent history, which is the most substantial content the previous version of this page was missing. In April 2015 demonstrations blocked access roads on Mauna Kea; about 300 protesters gathered and 23 arrests were made on 2 April; the governor halted construction for a week on 7 April. On 2 December 2015 the Hawaii Supreme Court invalidated the 2011 permit, holding that the Board of Land and Natural Resources had put the cart before the horse in approving it before the contested case hearing, without meaningful due process. On 28 September 2017 the board approved a revised conservation district use permit conditioned on removing three existing telescopes, and on 30 October 2018 the Hawaii Supreme Court upheld the revised permit by four votes to one.

Established Then July 2019, which is the part usually compressed into a clause. A renewed blockade began; 38 kupuna — elders — were arrested on 17 July; and a four-week blockade shut down all twelve observatories on Mauna Kea, the longest closure in fifty years. In July 2023 a new state-appointed authority began a five-year transition to assume management of the mountain, including Native Hawaiian community representatives and cultural practitioners, with full control transferring in July 2028. Frontier The statute creating that authority was not obtainable for this rewrite and the fetched accounts describe the body without naming the act, so the citation should be checked before the statute is quoted.

Established The land tenure fact frames all of it. The existing master lease expires in 2033, after which 40 of the 45 square kilometres revert to the State of Hawaii, and three facilities are scheduled for decommissioning under the comprehensive management plan. Established And the objection must be stated in its own terms rather than as an obstacle: Native Hawaiians hold that the entire summit is sacred and that developing the mountain — even for science — would spoil it. That is a position held on its own grounds.

Established The quotation from the other side of the same table is the more interesting one. The activist Noe Noe Wong-Wilson, on the new authority: “It's still early in the life of the new authority, but there's actually a pathway forward.” Frontier That is a consent process producing a possible yes, not a permanent no — and it is the strongest available evidence that consent at an observatory site is a constraint that can be satisfied rather than a veto.

Established The alternative site did not resolve anything, and why it did not is the transferable lesson. After the 2019 protests the telescope's international observatory sought permits at Roque de los Muchachos in the Canary Islands. On 29 July 2021 a Spanish judge revoked the 2017 land concession, holding that it lacked an international scientific treaty and questioning the observatory's firm commitment; environmental groups opposed construction in a protected refuge containing three Guanche archaeological sites. The legal cases were addressed by 2023, and on 23 July 2025 the Spanish government announced a €400 million investment to bring the telescope to the Canaries. Scientifically the site is lower, admitting water-vapour interference at mid-infrared wavelengths that bears directly on exoplanet and cosmology programmes, and a move would disrupt the existing partnership.

Frontier So the structural lesson generalises, and it is stronger than “siting is contentious.” The alternative site had its own consent problem — indigenous archaeology, environmental designation and a court ruling — with a different community and a different legal system. Relocating a telescope does not escape the consent question; it changes which community must consent.

Established Now the sponsor's decision, which is the controlled experiment. In August 2020 the US National Science Foundation began informal outreach to stakeholders, acknowledging delays in environmental review while seeking a more inclusive, meaningful and culturally appropriate process. In July 2022 it announced a new environmental survey covering both Mauna Kea and the Canary Islands, with continued funding contingent on the results, technical readiness updates and public comment. In December 2024 its External Evaluation Panel reported that site access was “an existential threat to TMT” and that mitigation via La Palma was insufficient. In June 2025 the foundation dropped support for the Thirty Meter Telescope in favour of the Giant Magellan Telescope; the international consortium stated it would continue its efforts.

Frontier That is a sponsor ruling against its own investment, and it should be weighted accordingly rather than reported neutrally. The foundation had sponsored the project's development, funded its design work, and run a multi-year process looking for a way to fund it. The agency with every institutional reason to protect its own investment concluded that site access was existential and stopped. An adverse finding from a sponsor about its own project carries more weight than the same finding from a critic. Established The proximate constraint was budgetary: the foundation could advance only one US giant.

Frontier One dating question must be flagged rather than smoothed. The previous version of this page attributed the decision to the 2026 budget request, while the retrieval path for this rewrite dates the drop to June 2025 and the panel finding to December 2024. Those are consistent if the June 2025 decision was expressed in the FY2026 request, which is the most likely reading. Neither date is asserted here as a citation: the foundation's own statement and the National Science Board record were not reachable for this rewrite, and the sequence — December 2024 panel, 2025 decision, FY2026 request — is what this page states.

Frontier The comparison that makes it an experiment is worth laying out flatly. The two projects share a generation, a segmented-mirror technology, an adaptive-optics lineage, international consortia, completed designs and published cost estimates of about $1.0 billion against about $1.4 billion. The Atacama site produced no comparable consent dispute; Mauna Kea did. One has all seven mirrors cast and a mount in manufacture; the other has a revoked concession on its backup site and a €400 million offer it has not taken up. The independent variable was not engineering.

Established The space line's near-term status, corrected. Roman launched at 07:26 EDT on 30 August 2026 and is on a three-month journey to its final orbit, with first images expected in early 2027; its field of view is at least 100 times that of Hubble, it will measure light from roughly a billion galaxies and conduct a statistical census of planetary systems, and it carries a coronagraph intended to block starlight in order to see exoplanets and planet-forming disks directly. Frontier Any sentence about Roman in the future tense is now wrong, and the previous version of this page contained several.

Established And the flagship behind it, with the budget line that is its clearest fact. The Habitable Worlds Observatory was recommended by the Astro2020 decadal survey, sits under a Great Observatory Maturation Program established in 2023, and had industry proposals selected on 5 January 2026 to advance its technologies. Planned aperture is 6 to 8 metres, with larger designs possible depending on 2040s launch vehicles; launch is given as the 2040s and cost as at least $11 billion. The FY2027 budget proposal requests $5 million against $150 million enacted for FY2026 — a reduction of about 97%, which is arithmetic on the two figures.

Established The physics that flagship must beat is a contrast ratio. Imaging an Earth-like planet around a Sun-like star requires suppressing the star by one part in ten billion at small angular separation. Frontier What has been demonstrated against that requirement — in the laboratory or on sky — was not obtainable for this rewrite, and no demonstrated-contrast figure appears on this page. That gap is the single most load-bearing number in the whole case for the mission, and a page that quoted the requirement without noting that the achievement is unsourced would be misleading by omission.

Established The open questions here are unusual for this category in that almost none of them are about physics. They are about schedules, architectures, appropriations and consent, plus two genuine instrument questions.

Frontier Question one: do the ground giants hit their current dates? The European project states technical first light in March 2029 and science operations in December 2030; the same organisation's 2011 plan said 2024–2025. The Giant Magellan consortium states commissioning in the early 2030s, with three of seven primaries still in polishing as of 2023. Both dates come from the bodies building the telescopes, and the previous dates from those same bodies slipped by about five years.

Frontier Question two, and it is a live engineering disagreement rather than a settled matter: many small segments or few large ones? The European telescope uses 798 segments at 1.45 metres; the halted 30-metre design uses 492; the Giant Magellan uses seven at 8.417 metres. Many small segments means mass production, many actuators and complex phasing. Few large ones means each is a multi-year casting and polishing campaign with far simpler control. Neither approach is obviously right, and this page presents them as an open question rather than treating segment count as a proxy for progress.

Frontier Question three: how far does adaptive optics scale? Correction complexity grows with the number of atmospheric coherence cells across the aperture, which grows steeply with diameter and steeply with decreasing wavelength. Frontier No scaling exponent, isoplanatic angle or cone-effect magnitude appears on this page, because the fetched source explicitly lacks all three. The qualitative claim — complexity grows much faster than aperture — is secure and is the point; the formula is not asserted.

Frontier Question four: is site access an existential threat to the halted project, or a solved problem waiting on a process? The sponsor's own evaluation panel said existential in December 2024 and judged relocation insufficient mitigation. Against that: the consortium says it continues, Spain offered €400 million in July 2025, and a community representative on the new stewardship authority describes a pathway forward. Full control of the mountain does not transfer until July 2028 and the master lease expires in 2033.

Speculative Question five: does the next space flagship exist? It is pre-formulation, with no confirmed budget, no launch date, a planning placeholder in the 2040s, an estimate of at least $11 billion, and a 97% single-year cut in its FY2027 request. Everything about the science case is strong and everything about the programme is contingent.

Frontier Question six: do the giants together deliver a credible biosignature detection? This is the prize the whole field is organised around, and it is contingent on the flagship existing, which is contingent on a budget line currently requested at $5 million. The ground giants alone are a partial answer, because atmospheric transmission and contrast limits bite hardest exactly where an Earth analogue sits.

Speculative Question seven, kept short and kept flagged: the exotic end. A telescope placed in the solar gravitational lens focal region beginning around 542 to 550 AU would see light amplification up to about a hundred billion and angular resolution of order a ten-billionth of an arcsecond; the proposers suggest a 1-metre telescope there could resolve surface features on an exoplanet at 30 parsecs to roughly 10 kilometres, sufficient for continents, oceans and seasonal change, reached by solar sails in about seventeen years. The authors' own framing is that such a mission is expected to be a difficult task. Speculative Lunar far-side radio arrays and large space interferometers are genuine concepts, decades away, and not part of the current build-out.

Handwave And the framing itself. “Ever-larger telescopes follow from ever-better engineering” fails on its own evidence: two technically comparable US giants entered the decade and the one that was dropped was dropped over site access and a budget that could carry only one. The engineering was the part that went well in both cases.

Established The first bottleneck is the cost exponent, and it is the reason this subject has a ceiling. Cost scales as roughly the 2.7 power of diameter, so doubling the aperture costs about six and a half times as much. Frontier A 100-metre ground telescope would therefore cost roughly fourteen times the 39-metre one — arithmetic on the published exponent — which is why the 100-metre concept was abandoned in favour of a 39-metre design, and why the framing is false as economics even where it is true as physics.

Frontier The second is correction complexity, which grows faster than aperture and is not a funding problem. The number of correction elements tracks the number of atmospheric coherence cells across the mirror, which rises steeply with diameter and steeply toward shorter wavelengths. The European telescope's answer is a 2.4-metre deformable mirror with up to 5,000 actuators running at a kilohertz, fed by six laser guide stars. Uncorrected, an 8-to-10-metre telescope delivers about one arcsecond; corrected, 30 to 60 milliarcseconds in the infrared.

Established Third, sky coverage, which the previous version of this page did not name at all. Adaptive optics needs a natural guide star of roughly magnitude 12 to 15 very near the target, and the source is blunt that this severely limits the application of the technique. Laser guide stars relieve the constraint without eliminating it, because tip-tilt correction still requires a natural reference. A telescope that works superbly on a fraction of the sky is a different instrument from one that works everywhere.

Established Fourth, segment count as a manufacturing reality. Segment number grows as the square of diameter, and each segment is an independently figured, mounted, phased and maintained optical component. 798 of them at 1.45 metres, or seven at 8.417 metres cast over more than a decade with three still in polishing as of 2023 — two different bets, both expensive, neither obviously right.

Established Fifth, and for the space line it dominates everything: starlight suppression. An Earth-like planet beside a Sun-like star requires a contrast of one part in ten billion at small separation. Frontier The gap between that requirement and demonstrated performance is the most important quantity in the case for the next flagship, and this rewrite could not source the demonstrated figure. The requirement is stated; the achievement is not, and the omission is deliberate.

Established Sixth, and it is the one that actually decides outcomes: a decade of unbroken appropriations. The completed infrared flagship overran roughly tenfold, slipped fourteen years, survived a cancellation vote and finished. The next one is requested at $5 million after $150 million enacted the year before. Frontier Neither of those numbers is an engineering fact, and both determine what gets built.

Established And seventh, consent at the site, which is not a risk to be mitigated but a condition to be met. Two Hawaii Supreme Court decisions going in opposite directions, 23 arrests in 2015, 38 elders arrested in 2019, a four-week shutdown of every observatory on the mountain, a state authority taking full control in July 2028, a master lease expiring in 2033 — and a revoked concession, environmental designation and indigenous archaeology at the alternative site, in another country and another legal system.

Established This brief records no dependency on another brief, and that is a substantive statement about the subject. Nothing here waits on a physics result or a demonstration that another page produces. Mirrors get cast, segments get polished, deformable mirrors run at a kilohertz, coronagraphs fly. What the giants wait on is appropriations and consent.

Established The first standing requirement is agency funding sustained across decade-long construction. The completed infrared flagship took roughly ten times its original estimate and fourteen extra years, and survived a committee vote to cancel it in July 2011 that ended with Congress capping completion funding at $8 billion. That is the observed distribution for a first-of-kind space observatory that worked, and the requirement is a political system able to absorb it.

Established The second is community consent at the observatory site. It is stated as an institutional requirement rather than a project risk because relocating does not escape it — it changes which community must consent, and the alternative site's own concession was revoked by a court in 2021 over environmental designation and indigenous archaeology.

Frontier One dependency is real, unowned by this page, and worth naming precisely. The next space flagship's architecture assumes serviceability — a mountaintop observatory at the second Lagrange point whose instruments can be upgraded. Space-Based Manufacturing owns whether on-orbit assembly and servicing exists; this page owns whether the design assumes it. No typed edge is claimed because the mission is pre-formulation and the assumption is not yet a requirement.

Frontier And two hand-offs in the science direction. Exoplanet Colonization owns the targets and what a biosignature would mean; this page owns the instruments and whether they get built, and states the one-part-in-ten-billion contrast as an engineering specification rather than as an interpretive question. Frontier Interstellar Probes owns getting a spacecraft to 550 AU; this page owns the lens as an instrument, for one paragraph.

Established The completed experiment that anchors the whole subject is the infrared flagship at the second Lagrange point. A 6.5-metre primary of 18 gold-coated beryllium segments, deployed after launch on 25 December 2021, working since. It demonstrates that segmented, deployable, cryogenic space optics work, which was the single largest technical risk anyone had identified.

Established The newest experiment launched days before this page was written. Roman lifted off at 07:26 EDT on 30 August 2026 with a field of view at least 100 times Hubble's and a coronagraph aboard, on a three-month cruise with first images expected in early 2027. Frontier Its coronagraph is the flight test of the technology the next flagship's case depends on, and the results of that test do not exist yet.

Established The ground experiments are the telescopes themselves and their first-light dates are the milestones. Technical first light for the 39-metre in March 2029, science operations in December 2030, commissioning of the 25.4-metre in the early 2030s. Both dates are the builders' own and both projects have already slipped against earlier versions of the same claim.

Established And one ground experiment already finished on schedule enough to be a counterexample. The Vera C. Rubin Observatory released first-light images on 23 June 2025 and began full survey operations on 30 June 2026, at a construction cost of about $680 million. Frontier A large, complex, ground-based observatory completed and operating is the fair counterweight to the two giants' schedule record, and it belongs on this page for exactly that reason.

Frontier The experiment this page most wants and could not source is a contrast measurement. One part in ten billion is the requirement for an Earth analogue. What has been achieved in the laboratory and on sky was not obtainable for this rewrite, and until that number is in hand the flagship's technical case cannot be evaluated by a reader of this page.

Frontier And an institutional experiment ran to completion in this decade, whether or not anyone designed it as one. Two comparable giants, one sponsor, one budget that could advance one of them, a multi-year process explicitly looking for a way to fund both sites, an external evaluation panel, and a decision. The result is on the record and it is about consent and money.

Established The primary-mirror requirement divides into two answered questions and one unanswered one. Segments can be many and small — 798 at 1.45 metres and 50 millimetres thick — or few and large, seven at 8.417 metres. Both are in fabrication today and both work; what nobody has demonstrated is which one is right at 30 metres and above, because only one telescope of each kind is being built.

Established The adaptive-optics requirement is a control system with a hard cadence. Wavefront distortions must be measured on a few-millisecond cadence, which sets detector readout and control-loop requirements that scale with actuator count. The 39-metre's chain is a 2.4-metre deformable quaternary mirror with up to 5,000 actuators at a kilohertz, fed by six laser guide stars. The payoff is 30 to 60 milliarcseconds in the infrared against about one arcsecond uncorrected, which is the entire justification for building a 39-metre mirror on the ground rather than a smaller one in space.

Frontier The guide-star requirement is the constraint that limits where the payoff applies. A natural reference of magnitude 12 to 15 must lie very near the target; laser guide stars relieve this but tip-tilt still needs a natural reference. Sky coverage is therefore a specification, not a detail, and the fetched source's own word for the limitation is “severely.”

Established The space requirement is thermal and mechanical stability more than aperture. A deployable segmented cryogenic primary was the hard problem for the current flagship and it was solved. For the next one the hard problem moves to the coronagraph: suppressing a star by one part in ten billion at small angular separation, held stable over an observation. Frontier The demonstrated performance against that requirement is not stated here because it was not obtainable, and the honest engineering statement is that the requirement is specified and the achievement is unsourced on this page.

Frontier The serviceability requirement is a design philosophy with a dependency attached. A 6-to-8-metre observatory intended to operate for decades with upgradable instruments assumes an on-orbit servicing capability whose existence is another page's question. Designing for servicing that does not yet exist is a bet, and it is worth naming as one.

Speculative And the exotic requirement, stated as what it would need. A solar gravitational lens instrument needs to reach a focal region beginning around 542 to 550 AU — proposed via solar sails of roughly sixteen vanes at a thousand square metres each, reaching about 150 km/s at perihelion and arriving in about seventeen years — and then to hold station along a focal line while integrating. The amplification and resolution figures are extraordinary and the transit is the problem, which is why Interstellar Probes owns it.

Established The seam that most needs drawing is with the exoplanet science page, and it runs on instruments against interpretation. Exoplanet Colonization owns the targets and what a biosignature would mean. This page owns the instruments and whether they get built, which is why the one-part-in-ten-billion contrast appears here as an engineering specification and the question of what a detection would license appears there.

Frontier The second is with communications, and it is concrete rather than thematic. Deep Space Communications shares optical and precision-pointing technology — and the 5-metre Hale Telescope at Palomar served as the downlink receiver for the deep-space laser demonstration, which is an astronomical aperture doing communications work. If optical deep-space service scales, it competes for the same class of aperture and the same class of site.

Frontier The third is the capacity seam, and it runs the other way from the usual one. Deep Space Infrastructure records that contention during one uncrewed test flight produced major impacts on flagship robotic missions including a space telescope. The instrument on this page is the thing that gets degraded when the network is short — a useful reminder that infrastructure scarcity is felt by science missions rather than by the infrastructure.

Frontier The fourth is assembly and servicing. Space-Based Manufacturing owns whether on-orbit assembly and servicing exists as a capability; this page owns whether the next flagship's architecture assumes it. Orbital Shipyards owns assembly as a facility. An observatory designed to be upgraded in orbit is a customer for both.

Frontier Fifth, satellite constellations, which this page mentions and does not develop. Interference with ground astronomy is partly a dark-sky observing constraint and partly a fact about the orbital population; the population belongs to the orbital-operations briefs and the observing constraint is noted here in the ethics section.

Speculative And sixth, the lens. Interstellar Probes owns getting to 550 AU; this page owns the lens as an instrument — amplification, resolution, what it would image — in one flagged paragraph; and Deep Space Communications owns at most one sentence on focal-region amplification as a link concept.

Established The institutional finding on this page is the strongest kind of evidence available anywhere in this cluster: a sponsor ruling against its own investment. The US National Science Foundation had sponsored the Thirty Meter Telescope's development, funded design work, and run a multi-year environmental and consultation process explicitly looking for a way forward on either of two sites. Its own External Evaluation Panel reported in December 2024 that site access was an existential threat and that relocation was insufficient mitigation, and in June 2025 the foundation dropped the project in favour of the Giant Magellan Telescope.

Frontier Interest runs against that finding, which is why it should be weighted up rather than reported neutrally. An agency has every institutional reason to protect a project it has already funded. When it concludes instead that the project is unbuildable, the conclusion carries more than the same words from a critic would. The proximate constraint was budgetary — only one US giant could be advanced — and the discriminator between two comparable designs was the site.

Established The consent institution on the mountain is now being rebuilt, and the timetable matters. A state-appointed authority began a five-year transition in July 2023, including Native Hawaiian community representatives and cultural practitioners, with full control transferring in July 2028; the master lease expires in 2033, when 40 of 45 square kilometres revert to the state; three facilities are scheduled for decommissioning. Frontier A community representative describes a pathway forward, which is a more informative datum than either side's advocacy.

Established The funding institution's behaviour is documented across a full project lifetime. A $1 billion, ten-year, 2007-launch concept became a $9.7 billion lifetime cost launching in December 2021, survived a July 2011 committee move to cancel, and was completed under an $8 billion congressional cap on remaining funding. Frontier The successor's FY2027 request of $5 million against $150 million enacted for FY2026 is the same institution behaving very differently, three years after a maturation programme was established and one year after industry technology contracts were let.

Frontier Consortium structure is the quiet variable in all of this. The European telescope is built by an intergovernmental organisation with member-state subscriptions; the Giant Magellan by a consortium; the halted project by an international observatory spanning several countries whose partnership a relocation would disrupt. A national funding decision can strand an international partnership, and that is what a $400-million-plus foreign offer to relocate is really about.

Established And the cluster-level observation, stated in full in Space Law and Governance. This brief carries constraints-only adjudications and no brief-to-brief edge. Mega-telescopes are not waiting on a discovery; they are waiting on a decade of appropriations and a community's agreement, and neither is produced by better engineering.

Established The sharpest ethical question in this subject is consent at a sacred site, and it must be stated in the terms the objectors use. Native Hawaiians hold that the entire summit of Mauna Kea is sacred and that developing it — even for science — would spoil it. That is a position on its own grounds, not an obstacle to a telescope, and a page that describes it only as a project risk has already taken a side.

Frontier The record also refuses the simplest counter-narrative. Two Hawaii Supreme Court decisions went in opposite directions; a permit was granted in 2017 and upheld four to one in 2018; a separate court in Spain revoked the alternative site's concession in 2021; environmental groups objected in the Canaries over a protected refuge and three Guanche archaeological sites; and a federal funding decision in the United States settled the outcome. Attributing the result to any single party misdescribes a multi-jurisdictional process.

Established The stewardship question has a date attached and that is unusual. A new authority including community representatives and cultural practitioners takes full control of the mountain in July 2028; the master lease expires in 2033 with most of the land reverting to the state; three telescopes are scheduled for decommissioning. Consent here is not a one-off approval but a tenancy with an expiry.

Frontier Access equity is the next question and it is structural. A handful of instruments costing one to two billion dollars each will define observational astronomy for a generation, and access is allocated by consortium membership and national subscription. A country that is not a member of one of three or four organisations does not observe with a 39-metre telescope, which is a distributional fact about knowledge production rather than about money alone.

Frontier Dark skies are the ethical question that arrives from outside the field entirely. Satellite constellations cross the same sky these instruments observe, and the constraint is real for wide-field survey work in particular. The orbital population belongs to the orbital-operations briefs; this page notes the interference and does not develop it.

Established And the opportunity-cost question, stated with its numbers. A flagship at $9.7 billion lifetime, a successor estimated at $11 billion or more, ground giants at $1 to $2 billion each. Frontier Whether that is the right allocation against other science is a judgement this page does not make, but it makes it in full view: the figures are on the page, and the cost exponent explains why they cannot be much smaller.

Established These are among the most powerful instruments the species has built, and the science case does not need inflating. A 39-metre primary corrected at a kilohertz, a 25.4-metre with 368 square metres of collecting area, a 6.5-metre infrared observatory already working at the second Lagrange point, and a wide-field survey telescope launched on 30 August 2026 with a field of view a hundred times Hubble's.

Frontier The civilizational prize is specific and it is contingent. Together with a space flagship able to suppress starlight by one part in ten billion, these instruments could deliver the first credible detection of biosignatures on an Earth-like world. That flagship is pre-formulation, has no confirmed budget or launch date, and was requested at $5 million for FY2027 after $150 million enacted for FY2026.

Established The deflating structural point is the exponent. Cost as roughly the 2.7 power of diameter means each generation of aperture costs about six and a half times the last for a doubling, which sets a ceiling that no manufacturing improvement has yet moved. “Ever-larger” is not a trend line; it is a sequence of increasingly expensive political decisions.

Frontier And the deflating institutional point is that the engineering is the part that is going well. Mirrors get cast. Segments get polished. Deformable mirrors run at a kilohertz. A deployable cryogenic segmented telescope unfolded correctly at a million and a half kilometres from Earth. Roman flew on 30 August 2026. What determines which giants exist is a decade of unbroken appropriations and a community that agrees to the ground being used.

Established The sentence to leave a reader with. Two US giants of the same generation, the same segment technology and comparable cost entered this decade; one is under construction and one is halted; and the finding that stopped it came from its own sponsor and was about site access rather than optics. The independent variable was not engineering, and that is the most useful thing this subject has to teach.

These horizons track three separate clocks — construction, appropriation and consent — because the framing under test assumes only the first one exists:

• 10 yr: Frontier Expect technical first light for the 39-metre in March 2029 and science operations in December 2030 on the builder's current schedule, whose predecessor said 2024–2025. Frontier Expect the Giant Magellan to commission in the early 2030s with the last of its seven primaries out of polishing. Established Roman returns first images in early 2027, and its coronagraph results are the nearest thing to a test of the next flagship's core technology. Frontier Full control of Mauna Kea transfers to the new authority in July 2028 and the master lease expires in 2033, both inside this window.
• 25 yr: Frontier This is the window in which the ground giants deliver their exoplanet and cosmology programmes, and in which the halted 30-metre either finds a site and a sponsor or does not. Speculative It is also the window in which the next space flagship either enters formulation with a real budget or joins the list of decadal recommendations that were never built; a $5 million request against $150 million enacted is not yet an answer, but it is data. Frontier A first credible biosignature claim is possible in this window and would be contested for most of it.
• 50 yr: Speculative A 50-to-100-metre ground telescope is physically conceivable and, at the published cost exponent, would cost several times the largest instrument ever built — which is why the concept that existed in the early 2000s was replaced by a 39-metre rather than deferred. Speculative Large space interferometers and lunar far-side radio arrays are the coherent long-horizon architectures and are concepts today. Frontier The consent question does not go away at this range; it becomes a question about the Moon.
• 100 / 250+ yr: Speculative A telescope in the solar gravitational lens focal region beyond 542 AU is the only concept in the record that changes what an image of another planet means — surface features at kilometre scale rather than a spectrum from a point — and its authors call it a difficult task. Handwave Beyond that the forecasting stops being useful: what this subject's history predicts at long range is not a technology but a pattern, in which the instruments that get built are the ones whose sponsors stayed solvent and whose sites stayed available.

• Depends on This brief records no dependency on another brief. Nothing in the subject waits on a physics result: segments are being cast and polished, deformable mirrors run at a kilohertz, a deployable cryogenic segmented primary unfolded correctly in 2022, and a coronagraph launched on 30 August 2026. The three physical limits that do bite are all understood — cost scaling as roughly the 2.7 power of diameter, correction complexity growing faster than aperture, and segment count growing as the square of diameter — and none of them is waiting on a discovery. One architectural assumption does point outward: a serviceable flagship at the second Lagrange point assumes an on-orbit servicing capability that Space-Based Manufacturing owns, and no typed edge is claimed for it because the mission is pre-formulation.
• Requires (not on this map) Both constraints are institutional, and the evidence for each is a completed case rather than a worry. For funding: the current infrared flagship was conceived at $1 billion for a 2007 launch, cost about $9.7 billion in lifetime terms, launched in December 2021, and survived a July 2011 committee move to cancel that ended with Congress capping remaining funding at $8 billion. It works. The requirement it demonstrates is a political system willing to absorb a tenfold overrun and a near-death vote across four administrations — and its intended successor, ranked first by the last decadal survey and given industry technology contracts in January 2026, was funded at $150 million in FY2026 and requested at $5 million for FY2027. For consent: the discriminator between two technically comparable US giants was site access. The sponsor's own External Evaluation Panel found in December 2024 that site access was an existential threat and that relocation was insufficient mitigation, and the project was dropped in June 2025 in favour of the one in the Atacama. The alternative site did not solve it either — a Spanish court revoked the concession in July 2021 over environmental designation and indigenous archaeology — which is why the constraint is general: relocating a telescope changes which community must consent, and does not remove the requirement.
• Enables No typed enabling edge is claimed, because the instruments enable a science programme rather than another brief's technology. The concrete relationship worth naming is with Exoplanet Colonization, which depends on these instruments for every target it discusses: the one-part-in-ten-billion contrast requirement stated here as engineering is the precondition for the biosignature question stated there.
• Adjacent Exoplanet Colonization owns the targets and the meaning of a detection; this page owns the instruments and whether they get built. Interstellar Probes owns reaching the solar gravitational lens focal region; this page owns the lens as an instrument. Deep Space Communications shares optical and precision-pointing technology, and the 5-metre Hale Telescope served as its deep-space demonstration's receiver. Deep Space Infrastructure documents the network contention that degraded a flagship observatory's operations. Space-Based Manufacturing and Orbital Shipyards own assembly and servicing. Outside this map: optical fabrication, control theory, atmospheric physics, and the appropriations process.

Handwave “Ever-larger telescopes follow from ever-better engineering.” Two US giants of the same generation, the same segmented-mirror technology and the same adaptive-optics lineage entered this decade with completed designs and published costs within about forty per cent of each other. One is under construction with all seven primaries cast; the other is halted, with a revoked concession on its backup site and its own sponsor's finding that site access was an existential threat. Established The independent variable was not engineering.

Established “Why not just build a 100-metre?” Because cost scales as roughly the 2.7 power of diameter, so doubling aperture costs about six and a half times as much. Frontier A 100-metre would cost roughly fourteen times the 39-metre on that exponent — which is arithmetic on a published heuristic, and which is exactly why the 100-metre concept was replaced by a 39-metre design rather than merely postponed.

Established “Roman is coming.” Roman launched at 07:26 EDT on 30 August 2026 and is on a three-month cruise, with first images expected in early 2027. Any sentence about it in the future tense is now wrong, including several in the previous version of this page.

Frontier “The next flagship is the next flagship.” It is a decadal recommendation in pre-formulation with a technology maturation programme, industry contracts awarded in January 2026, a 6-to-8-metre planned aperture, an estimate of at least $11 billion, a 2040s placeholder for launch, and a FY2027 request of $5 million against $150 million enacted for FY2026. Speculative Nothing about the science case is weak and nothing about the programme is confirmed.

Frontier “We will soon photograph continents on another Earth.” A space flagship would resolve an Earth-like planet as a point of light and read its spectrum; it would not image a surface. The concept that would image a surface — a telescope in the solar gravitational lens focal region beyond 542 AU, resolving features at roughly 10 kilometres on a planet 30 parsecs away — is described by its own proposers as a difficult task, and reaching the focal region takes about seventeen years after solving the propulsion problem.

Frontier “Coronagraph performance is nearly there.” The requirement is a contrast of one part in ten billion at small angular separation. What has actually been demonstrated in the laboratory or on sky was not obtainable for this rewrite, and no demonstrated figure appears anywhere on this page. Frontier The gap between those two numbers is the most load-bearing quantity in the flagship's case, and anyone making a readiness argument should supply it.

Frontier “Adaptive optics scales with a known formula.” Correction complexity grows with the number of atmospheric coherence cells across the aperture, steeply with diameter and steeply toward shorter wavelengths. No scaling exponent, isoplanatic angle or cone-effect figure appears here, because the fetched source explicitly lacks all three. The qualitative claim stands and the quantitative one is not manufactured.

Established “More segments means further along.” 798 segments at 1.45 metres, 492 at similar scale, and seven at 8.417 metres are three different bets on the same problem, not three points on one road. Many small segments means mass production and complex phasing; few large ones means multi-year casting campaigns and simpler control. Frontier Neither has been shown right at this scale, because only one telescope of each type is being built.

Frontier “The Thirty Meter Telescope is dead.” It is unfunded by its former sponsor, which is a different thing. The consortium states it continues; Spain offered €400 million in July 2025; and a community representative on the new stewardship authority describes a pathway forward. Frontier It is also true that its concession on the alternative site was revoked in 2021, that the site is scientifically inferior at mid-infrared wavelengths, and that its own sponsor called site access existential.

Established “Native Hawaiian opposition stopped the telescope.” That is not a complete account. The record contains two Hawaii Supreme Court decisions going in opposite directions, a permit granted in 2017 and upheld four to one in 2018, a Spanish court revocation on the alternative site, environmental objections in the Canaries, an evaluation panel finding, and a federal funding decision that could advance only one US giant. Attributing the outcome to one party misdescribes a multi-jurisdictional process.

Established “Mega-telescopes are speculative.” They are not, and the previous version of this page got that right. Two ground giants are in construction, a segmented cryogenic space telescope has been working since 2022, a wide-field survey observatory began full operations on 30 June 2026 at about $680 million, and Roman launched in August 2026. The uncertainty in this subject is about which projects survive their sponsors, not about whether the engineering works.