How stations enter the registry

OGS Tracker documents optical communication ground stations: sites built to close free-space optical links with spacecraft. This page sets out what qualifies a site for inclusion, how entries are sourced, and how corrections work. It is maintained independently by Evan Grey and free to use, with no login or paywall. Listing cannot be bought: operator verification is a paid service, and it never changes what a record says — only who has confirmed it.

Scope

A site is in scope when it is a ground terminal intended for optical space-to-ground or ground-to-space communication, operational, under construction, or credibly announced. This includes agency observatories adapted for optical comms, purpose-built commercial stations, and research testbeds that have demonstrated links.

Out of scope: satellite laser ranging (SLR) stations with no communications role, purely astronomical observatories, inter-satellite-link terminals, and quantum key distribution receivers with no classical comms capability, except where a single site does both, in which case the comms role is what's documented.

Retrofits count; retrofit potential does not. Low-cost focal-plane assemblies now convert an existing telescope into an optical ground station — DLR's SOFA is the clearest example — which makes “purely astronomical observatory” a description of a night's work rather than a permanent category. An observatory enters the registry from the moment a comms assembly is fitted and documented, and its entry covers the comms role only. It does not enter because it could be converted: the registry lists no candidate sites, and a telescope that might one day take a kit is not a station. Every entry records whether the station was purpose-built or retrofitted, onto what, and — where a source says so — whether the fitted capability is removable, because a site that is an OGS only on booked nights is not continuously available and should not be counted as though it were.

The retrofit test is the telescope, not the site. Of the 41 records carrying sourced build provenance, 33 are purpose-built and 8 are retrofits. Every retrofit here is a conversion of a telescope that already existed for something else: Palomar's 5.1 m Hale Telescope takes a deep-space downlink assembly between astronomy campaigns, and Matera and Lustbühel were satellite laser ranging stations before they were anything else. The reverse does not follow — standing at an old observatory is not evidence of a retrofit. The Observatoire de la Côte d'Azur hosts one of each: the MeO laser-ranging telescope was converted, while FROGS on the Calern plateau was built new. RAL Space's station at Chilbolton is purpose-built for the plainest reason of all — that observatory's instruments are radars, lidars and a LOFAR array, so there was no optical telescope on site to convert. The remaining 11 entries record no build type at all: that is a gap in the sources, not a quiet vote for purpose-built, and those stations are named on the outreach list rather than guessed at.

Inclusion criteria

Every entry needs at least a verifiable operator and a verifiable location from a published source. Beyond that, each carries a status tier that answers the first question anyone in the field asks: can I rely on this site today?

Sites that cannot be tied to a source stay out of the registry, no matter how widely they are rumored. Absence from the registry is a sourcing statement, not a judgment about a site's existence.

What each entry records

Every station is described across the same field groups, each cited to its source and left empty where undocumented:

• Site & sky, coordinates, altitude, and cloud-free line-of-sight where published.
• Hardware, build provenance (purpose-built, or retrofitted onto an existing telescope and which one), aperture, supply chain (each organisation that built the station, by role — telescope, mount, terminal, modem, adaptive optics, integrator, or in-house; aggregated on the supply chain page), adaptive optics, wavelength coverage, uplink beacon, and demonstrated uplink/downlink data rates.
• Standards & interoperability, the CCSDS standards a station speaks, the space terminals it has linked with, and the networks it schedules through.
• Access & availability, operational status, who can book passes, and the point of contact.
• Regulatory, licensing authority and status, ITU filing references, landing-rights posture, dual-use classification, and laser/aviation clearance, orientation only.

Announcements, not aspirations

The optical-comms field runs ahead of its own construction: press releases, MOUs and funding rounds routinely describe stations that don't yet exist. OGS Tracker records the announcement, not the aspiration. An announced or in-development site is listed only when a named operator commits to a named location, is stamped with the date of the claim, and is visibly flagged below operational status, never folded in with sites that have closed a link. Capability fields (aperture, manufacturer, wavelengths, demonstrated uplink/downlink rate) stay empty until documented, so a station's record can never look more built than it is. The demonstrated-link badge appears only where a specific closed space link is on record, it is the registry's honest line between a working station and a planned one. If a promised milestone slips, the entry keeps its original date and its status is walked back rather than quietly refreshed. Sites that are publicly claimed but don't yet clear this bar aren't hidden, they sit on the claimed / undocumented watchlist, each with the reason it falls short, and graduate to the registry only when a named operator, a named location, and a citable source all line up.

Weather

Cloud is the dominant availability constraint on an optical link, which makes it the field most likely to attract numbers nobody can check. The registry handles it two ways and never confuses them.

Per station, we record and never derive. Where an operator or a study publishes a cloud-free line-of-sight figure it is recorded with the basis it rests on, because operator claim and cited climatology study are not the same number and are never displayed as though they were. An unsourced percentage is treated as a blank. OGS Tracker computes no cloud statistics of its own and publishes no availability estimate for any individual site.

Across the registry, we compare what is possible with what is declared. What can be derived honestly from cited coordinates is which pairs of sites are positioned to back each other up: far enough apart that one weather system rarely covers both, near enough that a single pass crosses both horizons. The weather lens on the map draws every such pair, and marks the ones an operator has actually documented in site_sky.diversity_group. The distances behind it are stated assumptions printed beside the result, the same class of calculation as the 20° elevation floor behind the GEO reach figures. A drawn pair is a geometric possibility and never an arrangement: the registry does not know whether two operators have agreed, scheduled or tested a handover, and does not imply it. It is not a cloud model, and it is not a forecast.

The registry carries no live or forecast conditions. Current weather is an operations problem that belongs to whoever is scheduling the pass, and a feed of it would be the least auditable thing on a site whose whole claim is that every figure can be traced back to a source.

Market figures quoted on this site

The homepage quotes one forward-looking figure: 830+ stations required by 2035. That is a third-party industry projection, not registry data, and it is held to a different standard than a station record — it describes a modelled future, so it cannot be verified the way an aperture or a licensing authority can. It is quoted to frame why the registry exists, and is never used to fill a field in an entry.

The accompanying share — how much of that buildout is in place — is not quoted from anyone. It is computed on the page as documented operational stations divided by 830, and it currently rounds to about 5%. Earlier versions of this site carried a figure of roughly 10%, taken from industry commentary. We removed it: it was inconsistent with our own count, which is the more defensible number precisely because every station behind it is individually cited and can be audited by any reader. The registry can only ever undercount — stations we haven't documented exist — so treat 5% as a ceiling on how built-out the ground segment provably is, not as a claim about how few stations are physically standing.

Registry counts — documented sites, operational stations, countries, and this percentage — are computed live from the entries themselves and change only when a sourced record does.

Sourcing

Every field of every entry cites where it came from. Sources are ranked, and a higher-ranked source overrides a lower one:

• Primary, operator publications, regulatory and licensing filings, ITU records, peer-reviewed papers and conference proceedings (SPIE, ICSOS, IAC).
• Secondary, trade press, agency press releases, industry analyses.
• Direct, information provided by station operators, marked as operator-supplied and attributed with their consent.

Operator-supplied information ranks below published primary sources by design: an operator's say-so never overrides a filing or a paper. It can fill a gap, or trigger a re-check of the published record — it cannot beat one.

Fields with no adequate source are left empty rather than estimated. An empty field means "not yet documented," never "zero."

Regulatory fields

Regulatory annotations, licensing authority and status, ITU filing references, landing-rights posture, dual-use classification, are compiled from public filings and published law. They are provided for orientation, are simplified by design, and are not legal advice. Verify independently before relying on them for licensing, export-control or operational decisions. For how these break down country by country, see the regulatory landscape.

Operator verification

A station may carry a Verified operator seal. It means one thing: someone confirmed to speak for the operator — corresponding from the operator's own domain — has walked the record field by field with the data desk and confirmed it is correct. Verified records may also carry a direct booking button pointing to the operator's own scheduling page. Verification is a paid service for commercial stations (free for academic and government sites) and lasts twelve months; when it lapses the seal decays to a dated form and the booking button comes off. Operator corrections made during verification need sources and pass through the same review as anyone else's — the fee buys the review and the button, never the contents of a field.

Verified stations also get priority in the corrections queue. To be precise about what that means: priority changes how fast a sourced correction is applied, never whether it is. An unverified station's sourced correction is applied on the same rules — it may just wait behind the queue. Full terms on the for operators page.

Independence and funding

The registry is maintained by one named person, not an operator, network, or investor in any listed station. Its only revenue is operator verification fees, which fund the data desk. There is no advertising, no sponsored placement, no sale of reader data, and no paid tier that touches a record's contents, status, or position. Academic and government stations verify free precisely so the seal's population is set by review, not by ability to pay. If any of this changes, this page changes first — and the change will be visible in its dateline below.

Checking our work

You do not have to take the registry's word for anything on it:

• Every field cites its source — the bracketed numbers in a station record link to the source list, so any figure can be traced and re-checked.
• The dataset is exportable. The tracker's JSON export downloads the filtered records complete — every field, source, and citation; the CSV carries the headline fields for spreadsheets. Audit them, diff them, republish the facts.
• Every record is dated. The last-verified date is when the data desk last checked the entry against its sources; the operator seal carries its own, separate date.
• Station ids are never reused. A removed station's identifier is retired, so a citation to a record stays meaningful even after the record goes; status changes are walked back visibly, not quietly refreshed.

Corrections and removals

Corrections are free, for everyone, always — verification is never a toll on accuracy. Operators and researchers are invited to correct their entries; corrections that come with documentation are applied and the source updated, and operator-supplied corrections are marked as such. An operator may request that a contact field be removed; core existence data (name, operator, location, status) stays, since it is compiled from public sources.

To submit a correction or a missing station, use the contribute form on the tracker.

What this registry is not

• Not a booking service. It identifies operators and points to them — on verified records, via a booking button to the operator's own scheduling page. It does not schedule or broker passes itself.
• Not a completeness claim. It grows as documentation surfaces, best-documented sites first. A gap is a sourcing gap, not a verdict on whether a site exists.
• Not marketing. No operator pays for inclusion, placement, or a favorable status tier. The one paid service, operator verification, buys a review, a booking link, and speed in the corrections queue — never what a record says.
• Not legal advice. Regulatory fields are orientation only; see the note above.