How the atlas is built

01 · Scope

Inclusion rules

The catalog covers missions launched from Earth, crewed and uncrewed alike. Every record requires a stable mission identity, launch date, status, and at least one authoritative institutional source. Core records also require a celestial-body destination. Evidence-backed extended-scope observatories may instead study a non-body scientific domain and omit one; their primary operating regime must be non-geocentric, although Earth may be recorded when it is the scientific target.

The catalog is intentionally incomplete while the foundation is established. Absence is not a judgment about scientific importance, and uncertain optional measurements remain absent instead of being inferred.

02 · Evidence

Provenance

Sources are normalized records with an organization, title, URL, type, access date, and verification date. Missions, bodies, events, and future trajectory datasets refer to those source IDs so claims can be traced without copying citations between files.

Most records rest on NASA’s NSSDCA archive — more than every other publisher combined — with other NASA and ESA mission pages, JPL and NAIF technical records, and Jonathan McDowell’s GCAT behind it. Dates express only the precision stated by the source.

03 · Geometry

Trajectory quality

Availability and quality are separate, validated facts. A mission never receives a path merely because its route is commonly understood.

Precise

Generated from authoritative navigation or ephemeris data with explicit provenance.

Reconstructed

Derived from authoritative records, with documented processing and limitations.

Approximate

A simplified, sourced representation suitable for orientation, not navigation.

Unavailable

No verified trajectory dataset is committed, so the atlas renders no path.

04 · Coverage

What the catalog currently contains

The catalog currently contains 307 missions, 60 celestial bodies, and 645 sources. Every number on this page is generated from the committed records, never written by hand.

Missions with a committed trajectory

6 of 307

Precise trajectories

6

Reconstructed trajectories

0

Approximate trajectories

0

Metadata fully sourced

278

Metadata needing review

29

Sources with verbatim excerpts

643 of 645

Destination batches inventoried

14 of 14

Trajectory coverage deliberately lags metadata coverage: a mission is catalogued once its record is sourced, not once a path exists for it. Absence of a trajectory is a stated fact about the data, not a gap in the mission’s history.

Every batch has been inventoried, every committed mission is claimed by the batch it belongs to, and every fabrication a committed audit found has been corrected: the catalog’s enumeration has been audited. That is a narrower claim than “complete” — it does not require every enumerated candidate to be resolved (some may still be pending, see below), it does not require every audit to have been blind re-research rather than a mechanical claim-support check, and it says nothing about a mission no enumeration ever named.

3 enumerated candidates are still unresolved: launched missions that an enumeration named and that research has resolved to neither a committed record nor a cited exclusion. All current unresolved candidates are retained evidence limits, not unfinished segment work. Each one records what is missing and what was attempted. They are left open rather than excluded, because recording a real mission as out of scope would be the more misleading answer.

05 · Publishing

Static-first architecture

Human-readable JSON records are validated with strict schemas, then compiled into a lightweight, deterministic mission index. Pages are server-rendered from committed data; only directory filtering runs in the browser.

Validation, index-drift checks, tests, and a production build run before release. This keeps the published snapshot reproducible and avoids a runtime scientific API.