Credits and sources

The app

The app at solartimeline.org: the Film; the Library, the film beside its 3D sets and the spacecraft and stations part by part; and the Explorer, the solar system on any date, with real missions replayed and today's craft live. It uses all of the film's sources (below, under The film) and these of its own.

Moons

Asteroids, comets and the Sun

Spacecraft and missions

Spacecraft visits

• OneWeb first-generation reference prototype: original geometry and plain PBR materials in tools/assets/real_oneweb.py. Approximate 1 × 1 × 1.3 m bus and 6 m array span from Johnson, Scott & Thorsteinson, AMOS 2021. Layout reference: OneWeb Satellites illustration published by SpaceTech, 23 March 2021. Antenna complement: Arianespace ST29 mission briefing. No third-party mesh, image, logo or texture is redistributed. Reflector sizes, solar-cell divisions, structural details and surface finish are illustrative. Basic shape approved for reference-model integration; bounded catalogue coverage uses representative geometry, not verified serial-specific details. Not a model of the newer replacement generation.
• Starlink V2 Mini reference model: original geometry and plain PBR materials, built from the F9-2 bus and solar-array dimensions in SpaceX's 4 October 2022 FCC supplement, Exhibit B p. B-1 (primary filing mirrored by Jonathan McDowell), and SpaceX's Second Generation Starlink Satellites. No third-party mesh, photograph, texture or logo is included. Thickness, cell divisions, hinges, antenna patches and thruster placement are illustrative. Review prototype; the model is not a verified match for each NORAD object.
• Visits use the existing JPL Horizons trajectories and CelesTrak/SGP4 positions. Satellite neighbours are sampled at the same second; proximity does not establish a common launch. Orientations are illustrative. The generic stand-in is original schematic geometry, with no claimed dimensions for the named spacecraft. Earth's visit backdrop uses the credited NASA Blue Marble map.

The whole scene

Listening to the deep-space craft

Live: NASA's Deep Space Network Now (eyes.nasa.gov/dsn; its dsn.xml and config.xml, NASA/JPL-Caltech): which dish at Goldstone, Madrid and Canberra is hearing or commanding which craft, the signal's rate, band and strength, the craft's range; read in the browser every 10 seconds in Live mode. NASA data, public; credited NASA/JPL-Caltech.

Earth in the Explorer

The Explorer's Earth layers.

Maps, shapes, pictures and landings in the Explorer

Maps

The Explorer shows each map's credit beside its globe.

Layers (from 2026-10-22)

More layers on the worlds, each credited on screen while it is on. Not in a frame: move a row to "Earth in the Explorer" above if one ever is.

The Saturn-moon, Pluto and Charon shapes and rotations are the IAU 2015 values, checked against NASA NAIF's pck00011.tpc. Their masses and orbits, and the corrected Moon, Earth and gas-giant figures, are from NASA NSSDCA's planetary fact sheets (from the Internet Archive's 2024 copies).

Shape models

Irregular bodies are drawn from their measured shapes.

Non-commercial: the Rosetta OSIRIS shape models (67P, Lutetia, Steins) are ESA archive data, which ESA distributes under CC BY-NC 3.0 IGO (ESA Space Science Archives terms, checked 6 Oct 2026; the DSK files themselves state no licence). The 67P DSKs are DLR's SHAP4S, not the later SHAP7 of Preusker et al. (2017). They are used unchanged in shape apart from decimation; the converted models are adapted material, credited as above, and go out non-commercially only; any commercial use needs ESA's licence ([email protected]). There is no ShareAlike condition.

The bodies with pictures but no shape model are drawn as ellipsoids of their published size and said so: Borrelly (Deep Space 1 saw one side; the IAU's length and SBDB's mean diameter), Annefrank (Duxbury, T. C. et al. 2004, JGR 109:E02002), Dinkinesh and its moon Selam (Levison, H. F. et al. 2024, Nature 629:1015), Donaldjohanson (NASA, Lucy's first images, 21 Apr 2025), Dactyl (Belton, M. J. S. et al. 1996, Icarus 120:185). The interstellar visitors 1I/'Oumuamua, 2I/Borisov and 3I/ATLAS were never resolved: spheres of their estimated sizes ('Oumuamua ISSI Team 2019, Nature Astronomy 3:594; Jewitt, D. et al. 2020, ApJL 888:L23; Jewitt, D. et al. 2025, ApJL 990:L2). Tempel 1 turns at Deep Impact's rate before its 2011 perihelion and Stardust-NExT's after (the IAU's two epochs); Hartley 2 and Toutatis tumble and are drawn true at their flybys only; Wild 2, Borrelly, Halley, Mathilde and the rest have nominal turns, said in their fact sheets.

The new bodies' sizes and rotations are the IAU 2015 values (checked against pck00011.tpc with SPICE: under 0.0001 degrees), or the mission's own PCK where the IAU has none (Bennu, Ryugu, Didymos; Arrokoth from Porter et al. 2024). Masses are GM from NASA JPL Solar System Dynamics (planetary and satellite physical parameters, the Small-Body Database) divided by G.

Comet 67P landings

Philae's landing (12 November 2014) and Rosetta's end (30 September 2016) on 67P, as the Explorer shows them; each image's credit and licence is shown beside it.

The OSIRIS and NAVCAM frames are ESA's archive originals: OSIRIS frames under CC BY-SA 4.0 and NAVCAM frames under CC BY-SA 3.0 IGO, by ESA's own statements (the archive's 2018 copyright page, read from the Internet Archive, gives both credit lines).

ShareAlike: the web copies of the CC BY-SA images (resized; the NAVCAM frames stretched from 16 bits) are adapted material and go out under the same licence as their source (CC BY-SA 4.0 or 3.0 IGO) with the credit above and a link to the licence, not under the film's CC BY-NC.

Classic pictures of the planets and moons

First photograph of Earth from space (24 October 1946). V-2 No. 13, White Sands, New Mexico; U.S. Army / White Sands Missile Range / Applied Physics Laboratory. Army source and U.S. Army Photo credit, original image. Public domain in the United States as an official U.S. Army photograph. NASA historical context establishes 65 miles (about 105 km), 35 mm film and a frame every 1.5 seconds. Unaltered black-and-white source, resized for the web. Pin at approximately 32.4° N, 106.5° W represents the launch area, not a measured camera position; 105 km altitude and 60° viewing field provide illustrative context. Exact time, pointing and lens are not reconstructed; noon UTC is a date-navigation placeholder.

Classic spacecraft pictures, each pinned on its body where it was taken from and flown to from there: the times from the PDS Ring-Moon Systems Node's OPUS and the image pages, the camera's place from JPL Horizons (the PDS trajectory tables for Galileo at Gaspra and Ida, a published sub-camera point where Horizons has no craft), each range and field checked against the source page. Also newer pictures (2015 on): the Artemis I and II frames timed by their cameras' EXIF clocks from the NASA Image and Video Library, Orion placed from JPL Horizons; BepiColombo's under ESA's licence.

Mariner 4's pictures of Mars, 15 July 1965: frames 1, 6, 7, 8, 11, 15, 17, 18 and 19, the pair mosaics of frames 1-2, 3-4, 5-6 and 7-8, and the hand-coloured first picture, each pinned at the camera's place on the fitted hyperbola, looking at the frame's published centre; the times, ranges, centres and Sun angles from JPL's Mariner Mars 1964 Bulletin No. 52 (4 August 1965; the scan held by CSIRO Parkes).

Other landings and impacts

The Explorer's landing timeline also plays the other real landings and impacts with public trajectories. Only computed positions are published. Where a path is corrected for the picture, the Explorer says so ("adjusted") and how.

Rovers' drives: Curiosity, Perseverance and Ingenuity from NASA/JPL-Caltech's MMGIS traverse maps; Spirit and Opportunity from the PDS Geosciences Node MER Analyst's Notebook (NASA/JPL-Caltech/Washington University); Sojourner from the PDS rover uplink logs (NASA); Zhurong from Ding et al. 2022 (Nature Geoscience) and the RoPeR camera labels (figshare 10.6084/m9.figshare.29946872, CC BY 4.0, CNSA/GRAS); Yutu-2 from the Chang'e-4 radar data (Zenodo 10.5281/zenodo.10154147, CC BY 4.0, CNSA/GRAS); Yutu from Planmap D6.2 (Zenodo 10.5281/zenodo.4055234, CC BY 4.0) after Xiao et al. 2015.

Flybys, arrivals, entries and an impact (42 encounters at 27 bodies, each closest approach checked against the published figure): trajectories from NASA JPL Horizons (ssd.jpl.nasa.gov; Solar System Dynamics Group), except Rosetta at Steins and Lutetia from ESA's own kernels (ESA SPICE Service) and Galileo at Gaspra and Ida from the PDS one-minute trajectory tables (GO-A-POS-6-GASPRA/IDA-FLYBY-TRAJ-V1.0, NASA PDS PPI), and Hayabusa2's arrival at Ryugu from JAXA's home-position kernel (ISAS/JAXA). Turned into each body's frame with the Explorer's own rotation models. Artemis I and II whole and at the Moon: NASA/JSC navigation's trajectories in Horizons (Artemis I -1023, Post_TLI_Orion_AsFlown_20221213; Artemis II -1024, rev. 20 Apr 2026), checked against NASA's Artemis I blog (the outbound and return powered flybys, 81 and 80.6 miles up) and NASA's Artemis II pages (6,545 km over the Moon; the distance record); the timelines' marks from Horizons' headers and NASA's mission blogs. Both missions also in legs (Artemis I in five, Artemis II in four), joined where NASA said Orion crossed the Moon's sphere of influence (the Artemis I blog's flight days 5, 8, 18 and 21; Artemis II's from Horizons' header), checked against NASA's figures: the outbound flyby 130.4 km up at 12:57:37 UTC (NASA: 81 miles, 6:57 a.m. CST), the return flyby 129.5 km up at 16:43:58 (NASA: 80.6 miles, 10:43 a.m. CST), Artemis II 8,281.9 km from the Moon's centre at 23:00:46 (Horizons: 8,282 km at 23:01) and past Apollo 13's 400,171 km at 17:57 (NASA: 1:57 p.m. EDT). Hera past Mars and Deimos and at Didymos (Horizons -91: ESA's tracking, then its planned approach). The whole missions and Hera's arrival drawn fixed in space. CAPSTONE: its four-month ballistic lunar transfer and its first turns in the near-rectilinear halo orbit, from JPL Horizons -1176 (Advanced Space's orbit determination for NASA, tracked to 15 July 2026), checked against NASA's figures: farthest from Earth 1,538,000 km on 26 Aug 2022 (NASA: about 1.5 million km), the insertion at 00:39 UTC on 14 Nov 2022 (NASA: 7:39 p.m. EST), perilunes 1,630 to 1,760 km over the north pole and apolunes 70,400 to 71,600 km from the centre over the south (NASA: within 1,000 miles of one pole, 43,500 miles from the other). Gateway as designed (planned, never flown): NASA's reference 9:2 halo orbit, receding_horiz_3189_1burnApo_DiffCorr_15yr.bsp (D. E. Lee, NASA JSC; NAIF MORE_PROJECTS/DSG; Lee, D. E. (2019), White Paper: Gateway Destination Orbit Model, NASA NTRS 20190030294), with JPL's DE440 for the Moon; its status from NASA's Ignition fact sheet (24 Mar 2026). Planned sites near the Moon's south pole: NASA's nine Artemis candidate landing regions, traced from NASA's map (NASA, NASA Provides Update on Artemis III Moon Landing Regions, 28 Oct 2024; map NASA/GSFC/ASU/JSC/AGDT, public domain) and checked against the USGS Gazetteer of Planetary Nomenclature (IAU/USGS: the craters' centres); Moon Base I's target on the Shackleton connecting ridge (NASA, NASA Provides Update on Moon Base Rovers, Landers, Missions, 26 May 2026), placed on the ridge's highest ground in NASA's LOLA 20 m DEM (Barker et al. 2023) from LROC's junction with Shackleton's rim (Robinson, M., LROC, Traversing the Shackleton de Gerlache Ridge, 2022). Artemis III (planned): NASA's Artemis III mission page and the Planetary Society's account of its crew (June 2026). The film's supply run to Shackleton Base (fiction): flown by our own guidance over a point-mass Moon (DE440's GM) to the base's place on the LOLA 20 m DEM (Barker et al. 2023), the Sun from DE440. The comets and small bodies added 2026-10-22, each closest approach checked against the published figure: Deep Impact's impactor and flyby craft at Tempel 1 (di_finalenc_nav_v3, NASA PDS NAIF DI-C-SPICE-6-V1.0; A'Hearn, M. F. et al. 2005, Science 310:258), Stardust-NExT at Tempel 1 and Stardust at Wild 2 (sdu_l_2011_t1 and sdu_w2_opnav, NAIF SDU-C-SPICE-6-V1.0; Veverka, J. et al. 2013, Icarus 222:424; Brownlee, D. E. et al. 2004, Science 304:1764), EPOXI at Hartley 2 (dif_dixi_nav_v1, NAIF DIF-C/E/X-SPICE-6-V1.0; A'Hearn, M. F. et al. 2011, Science 332:1396), Deep Space 1 at Borrelly (ds1_opnav_borrelly, NAIF DS1-A/C-SPICE-6-V1.0; Soderblom, L. A. et al. 2002, Science 296:1087), Giotto at Halley (giotto_19860305_19860317, NAIF from ESOC's ephemeris, ESA SPICE Service; Reinhard, R. 1986, Nature 321:313), and from JPL Horizons Stardust at Annefrank (Duxbury et al. 2004), NEAR Shoemaker at Mathilde (Veverka, J. et al. 1997, Science 278:2109) and Lucy at Dinkinesh and Donaldjohanson (Levison et al. 2024; NASA).

Mariner 4 at Mars, 15 July 1965 (not in JPL Horizons): the areocentric hyperbola JPL fitted to the encounter tracking, JPL Technical Report 32-881, Mariner Mars 1964 Project Report: Mission Operations (1966), Table 4 (its misprinted a and e corrected from its own periapsis), drawn as a two-body conic about Mars with GM and the Sun from JPL DE440; checked against JPL's Mariner Mars 1964 Bulletin No. 52 (4 August 1965: each picture's time, range, centre and Sun angle), the occultation's entry and exit in TR 32-881 and in Kliore, A. et al., *Mariner IV Occultation Experiment* (JPL TR 32-984, 1966) and Science 149:1243 (1965), and the closest approach in NASA's PIA14032 caption; 1965 UTC from the USNO's TAI-UTC table. Mariner 2 at Venus, 14 December 1962: NASA JPL Horizons (target -2), John Anderson's 2012 fit to the encounter Doppler printed in JPL TR 32-816 (1967), as NAIF's notes on m2_620827_621231_ja_v1 say; the closest approach usually quoted (NSSDCA, 34,773 km at 19:59:28 UT) is the 1962 navigation's, 123 km lower and 4 min 44 s later than this path. NASA documents and data, public.

Apollo 11's whole trip in eight legs: MSFC, Saturn V Launch Vehicle Flight Evaluation Report AS-506 (MPR-SAT-FE-69-9, 1969); Northrop for NASA MSFC, AS-506 Post-Launch Operational Trajectory (NASA CR-61291, 1969); NASA MSC, Apollo 11 Mission Report (MSC-00171, 1971); F. V. Bennett, Apollo Lunar Descent and Ascent Trajectories (NASA TM X-58040, 1970); the Apollo Lunar Surface Journal (E. M. Jones) for the landing's callouts; the Apollo Flight Journal's copy of the post-launch report; LROC for Tranquility Base; JPL DE440 for the coasts' gravity. NASA documents, public.

Its hardware by stage on the follow camera: the gallery's Saturn V, LM, CSM and command module, the stack changing at the flight evaluation report's staging times (S-IC/S-II separation 162.3 s, S-II aft interstage 192.3 s, escape tower 197.9 s, S-II/S-IVB separation 549.0 s; Tables 4-1 to 4-3) and the Mission Report's (CSM separation 3:17:04.6, docking 3:24:03.1, LM extraction 4:16:59.1, CM/SM separation 194:49:12.7, drogues 195:12:06.9); the LM's landing gear folded until Aldrin's "Landing Gear Deploy, Fire" at 098:14:35 (the Apollo Flight Journal's onboard transcript); the drogue and main parachutes (our models, below) over the CM from the drogues' deployment and the nominal 10,000 ft. Adjusted: the LM's height in the adapter (its folded footpads just over the instrument unit) and the CSM 15 m from the LM while turning round are ours; the LM's probes are hidden while its gear is folded (the model's are folded flat, as landed); the attitudes are not the flown ones (the stack points along its motion, the CM's heat shield into it; the CM hangs upright under its parachutes, not at its hang angle).

The deep-space craft's whole journeys in the overview: NASA JPL Horizons (heliocentric vectors), with NASA's Voyager encounter trajectory files (NAIF) for the Voyagers' planetary flybys and ESA's reconstructed Rosetta cruise (ESA SPICE Service); the Voyagers' termination-shock and heliopause crossings as published. DART (JPL's final reconstruction, -135) to its impact on Dimorphos: Daly, R. T. et al. (2023), Nature 616:443, doi:10.1038/s41586-023-05810-5 (the impact); Thomas, C. A. et al. (2023), Nature 616:448, doi:10.1038/s41586-023-05805-2 (Dimorphos's orbit 33 ± 1 minutes shorter). Hera (-91: ESA's tracking, then its planned approach to Didymos): checked against ESA's Mars flyby (around 5000 km up) and Deimos (as close as 300 km) pages. The Deep Space Network's craft (2026-10-22), each from its flight team's trajectory in JPL Horizons: Psyche (JPL navigation), Europa Clipper (JPL navigation; NASA/JPL's Mars flyby, about 884 km up), Lucy (its navigation team's reference trajectory to 2033; NASA's Dinkinesh and Donaldjohanson flybys), Juice (ESA), Solar Orbiter (ESA), BepiColombo (ESA; ESA's Mercury flybys), Hayabusa2 (JAXA), STEREO-A (from 2022), ESCAPADE Blue and Gold (UC Berkeley SSL), MMX (JAXA), and the cruises to Mars of Mars Reconnaissance Orbiter, MAVEN, Mars Odyssey (JPL navigation) and the ExoMars Trace Gas Orbiter (ESA flight dynamics). Their orbits at Mars and LRO's at the Moon in each body's paths (tools/flybys.py, Horizons). The craft near Earth (SOHO, ACE, Wind, DSCOVR, IMAP, SWFO-L1, Carruthers, JWST, Roman, TESS, Chandra, XMM-Newton, MMS 1 to 4): Horizons vectors from Earth's centre, where the DSN layer's lines end.

Not played (no public trajectory to the ground): Hayabusa's touchdowns on Itokawa (2005; no kernel covers the second, the first is meaningless below 20 m), Hayabusa2's touchdowns on Ryugu (2019; the final descents are not in the public archive).

The Library's spacecraft and stations

Real spacecraft in the gallery, each from its public model (NASA 3D Resources unless stated) or modelled by us, sized to published figures; what is adjusted or ours is noted in each fact sheet.

• Orion (NASA/Lockheed Martin; service module ESA/Airbus): our model, from published figures: NASA's Artemis I press kit and Reference Guide (the crew module 5 m across and 3.3 m tall; the wings' three 2 m panels, 7 m and 19 m), the Artemis II press kit, ESA's European Service Module pages (4.1 m across, its engines, radiators and tanks), Bibb, K. L. et al. (2011), AIAA 2011-3507 (the capsule's mould line) and JPL Horizons' headers (the 7.92 m stack); NASA's 3D-printing capsule (NASA 3D Resources) read for proportions only; finishes after NASA's flight photographs (the solar-wing camera selfies art002e004357, cmasaw1_20221118204608_034, art001e000415, art001e001619 and art002e031476, on which its markings, windows, thrusters and panels were placed and measured); the back shell's silver coating over its 1,300 tiles from NASA, "Engineers Refine Thermal Protection System for Orion's Next Mission" (19 Nov 2015, https://www.nasa.gov/missions/artemis/orion/engineers-refine-thermal-protection-system-for-orions-next-mission); NASA's insignia, logotype and the US flag drawn by us as painted on the flown craft (docs/licences.md). Reference photos NASA (public domain) and ESA (ESA Standard Licence), used for reference only.
• Hera (ESA): our model, from published figures: ESA's Hera launch kit (September 2024: the 1.6 m box, two 5 m wings of three panels, 14 m², the 1.13 m dish, the thrusters, the asteroid deck's instruments; the 10 N thrusters' 35 mm exits and area ratio 150 as ArianeGroup's 10 N bipropellant family, space-propulsion.com), DLR's Hera spacecraft page and ESA's photographs of the flight spacecraft (its blankets); no public 3D model exists. Reference photos ESA (ESA Standard Licence), used for reference only.
• DART (NASA/Johns Hopkins APL): our model, from published figures: Daly, R. T. et al. (2023), Nature 616:443 (the bus), APL (the 8.5 m roll-out arrays), NAIF's DART frame kernel dart_010.tf (layout), Bray (2023, the flat antenna), Shastry et al. (2017, NEXT-C), Fletcher et al. (2018, DRACO), APL's photographs; the MR-103G thrusters' area ratio 100 (satsearch's MR-103G listing); NASA 3D Resources' DART 3D-printing model for proportions only (not copied).
• CAPSTONE (NASA/Advanced Space; bus Tyvak): our model, from published figures: JPL Horizons -1176's notes (the 34 x 34 x 61 cm bus, 25 kg, two solar panel extensions, eight thrusters), eoPortal's CAPSTONE page (Stellar Exploration's propulsion, the X-band radio and S-band patch array), a 2.4 m span (keeptrack.space) and Tyvak's render (eoPortal, figure 2; reference only).
• Gateway (as designed; planned, never flown): NASA (NASA 3D Resources, Gateway Core), in metres as drawn (HALO 7.6 m with its docking rings and 3.3 m across its shielding, against Northrop Grumman's 7 by 3 m); its maps reduced for the web, the Government of Canada's wordmark on Canadarm3 painted out.
• Parker Solar Probe: NASA/Matthew J. Garcia (NASA 3D Resources, Parker Solar Probe), in metres as drawn (the heat shield 2.3 m across and the craft about 3 m tall, NASA's Parker Solar Probe launch press kit, August 2018); turned with its shield to the Sun.
• Mars Reconnaissance Orbiter: NASA/Brian E. Kumanchik and NASA/Christian A. Lopez (NASA 3D Resources, Mars Reconnaissance Orbiter (MRO) (C)), in metres as drawn (the 3 m dish and 5.35 m wings, NASA/JPL MRO launch press kit, 2005); SHARAD's dipole lengthened to its published 10 m.
• MAVEN: NASA/Brian E. Kumanchik (NASA 3D Resources, Mars Atmosphere and Volatile EvolutioN (MAVEN) (B)), in metres as drawn (the 2 m dish and 11.43 m length, NASA MAVEN press kit, 2013); the Langmuir Probe and Waves booms shortened to their published 7 m.
• 2001 Mars Odyssey: NASA/JPL/Eyes on the Solar System (NASA 3D Resources, Mars Odyssey), in metres as drawn (the 1.3 m dish, NASA/JPL 2001 Mars Odyssey launch press kit); the gamma ray spectrometer's boom lengthened to its published 6 m.
• Lunar Reconnaissance Orbiter: NASA/Brian E. Kumanchik and NASA/Christian A. Lopez (NASA 3D Resources, Lunar Reconnaissance Orbiter (B)), in metres as drawn (the array 4.27 x 3.20 m and the antenna 2.59 m out, NASA's LRO launch facts, 2009); its silver blankets brightened to NASA's photos of the flight craft (KSC-2009-2002).
• Psyche (NASA/JPL, ASU): our model, from published figures: NASA/JPL's Psyche press kit (2023: the 24.76 x 7.34 m span, the 2.2 x 2.4 m bus 4.9 m tall with its 2 m booms, two five-panel cross-shaped arrays of 11.3 x 7.3 m and 75 m², the 2 m dish, four Hall thrusters), NASA/JPL-Caltech's illustrations PIA21499 and PIA24473 (the cross layout) and NASA's photographs of the flight spacecraft (its finishes); NASA's engineering 3D model is no longer published.
• Europa Clipper: NASA Visualization Technology Applications and Development (VTAD) (NASA, Europa Clipper 3D Model), in metres as drawn (30.5 m across the arrays, NASA/JPL Europa Clipper press kit, 2024); simplified for the web.
• Lucy (NASA/SwRI/Lockheed Martin): our model, from published figures: NASA's Lucy launch press kit (September 2021: 15.82 m across, the 7.3 m UltraFlex arrays, the 2 m high-gain antenna, 3.95 m stowed, 2.00 m deep) and its Lockheed Martin render, Olkin, C. B. et al. (2021), PSJ 2:172 (the Instrument Pointing Platform and its two-axis gimbal), and NASA/KSC's photographs of the flight spacecraft; no public 3D model with a clear licence exists.
• Juice (ESA): our model, from published figures: ESA's Juice spacecraft specs and factsheet (the ten 2.5 x 3.5 m panels, 27 m tip to tip, the 2.5 m HGA, the 10.6 m boom, the 16 m RIME antenna, the RWI and Langmuir booms), the ESA SPICE Service's JUICE frames kernel juice_v46.tf and the Juice SOC's spacecraft-model table (where the parts sit; its CAD model measured, not copied), ESA's photographs of the flight spacecraft (its finishes; ESA Standard Licence, reference only).
• Solar Orbiter (ESA, with NASA): our model, from published figures: Müller, D. et al. (2020), A&A 642:A1 (its sizes, the heat shield's construction, and its renders for the layout), ESA's Solar Orbiter factsheet and heat-shield figures (3.1 x 2.4 m; the base 2.94 x 2.56 m), ESA's notes on the 1.1 m high-gain antenna, and NASA/KSC's photographs of the flight spacecraft; no 3D model with a clear licence exists (ESA's are under the ESA Standard Licence).
• ExoMars Trace Gas Orbiter (ESA/Roscosmos): our model, from published figures: ESA's Trace Gas Orbiter page (the 3.2 x 2 x 2 m body, 17.5 m wings, the 2.2 m HGA, the Electra helix), the ESA SPICE Service's Cosmographia model ExomarsTGO_v0.5 (where the parts sit; measured, not copied), ESA's photographs and renders (its finishes and instruments; ESA Standard Licence, reference only).
• Chandra X-ray Observatory: NASA/Brian E. Kumanchik (NASA 3D Resources, Chandra X-ray Observatory), in metres as drawn (13.8 m long, 19.5 m across the arrays: Chandra X-ray Center, Smithsonian Astrophysical Observatory); its silver blanket drawn as an even silver.
• Spitzer Space Telescope: NASA/Janice C. Lee and NASA/Robert L. Hurt (NASA 3D Resources, Spitzer Space Telescope), in metres as drawn (4.45 m tall, 2.1 m across: NASA Spitzer launch press kit, 2003); its normals cleared.
• Kepler: NASA/Brian E. Kumanchik and NASA/Christian A. Lopez (NASA 3D Resources, Kepler (A)), in metres as drawn (2.7 m across: NASA Kepler launch press kit, 2009; 0.3 m taller than its 4.7 m, left); its bus's silver blanket drawn as an even silver.
• TESS: NASA/Matthew J. Garcia (NASA 3D Resources, Transiting Exoplanet Survey Satellite (TESS) (B)), scaled by 0.5 to 3.7 m across the arrays (NASA/MIT; Ricker et al. 2015); its normal maps rewired as normal maps.
• SOHO (ESA and NASA): our model, from published figures: ESA's SOHO spacecraft page (4.3 m along the Sun-pointing axis, 9.5 m across the panels), the NASA/ESA SOHO site (4.3 x 2.7 x 3.7 m), ESA Bulletin 84 (the service and payload modules, the 0.8 m high-gain antenna), and the NASA/ESA SOHO gallery's photographs of the flight spacecraft (its layout and blankets; reference only); NASA 3D Resources' SOHO model does not match these figures and is not used.
• STEREO (NASA, built by Johns Hopkins APL): NASA 3D Resources, Solar TErrestrial RElations Observatory (STEREO) (no modeller named), scaled to the 6.5 m span of NASA's STEREO mission booklet (2006); its high-gain dish (1.2 m), IMPACT boom (6 m) and S/WAVES antennas (6 m) corrected to the booklet; its blankets silver after the STEREO Science Center's photographs of the flight craft; its twelve thrusters ours.
• Mars Express (ESA): our model, from published figures: ESA's Mars Express spacecraft and engineering pages (the 1.5 x 1.8 x 1.4 m bus, 11.42 m² of cells, the 1.6 m high-gain antenna, the 400 N engine), ESA's MARSIS figures (two 20 m dipole elements, a 7 m monopole), the ESA SPICE Service's Mars Express frames kernel and shape models (where the parts sit; measured, not copied) and ESA's renders (its finishes; ESA Standard Licence, reference only).
• Hope, the Emirates Mars Mission (MBRSC, with LASP, ASU and UC Berkeley): our model, from published figures: Amiri, H. E. S. et al. (2022), Space Sci. Rev. 218:4 (CC BY 4.0: the 1.85 m high-gain antenna, the thrusters, the stowed size, its CAD drawings and photograph for the layout), MBRSC's figures (2.37 m wide, 2.90 m tall, 7.9 m across the arrays) and photographs of the full-size mock-up at IAC 2021 (Wikimedia Commons, reference only).
• BepiColombo (ESA/JAXA): our model of the cruise stack, from published figures: ESA's BepiColombo factsheet (the 6.3 m stack, about 30 m across the transfer module's wings, the MPO 2.4 x 2.2 x 1.7 m with its 3.7 m radiator, Mio 1.8 x 1.1 m), Benkhoff, J. et al. (2021), Space Sci. Rev. 217:90 (the four modules, the arrays, the radiator, Mio's panels, the MOSIF's cone; CC BY 4.0), ESA's T6 thruster notes, the ESA SPICE Service's Cosmographia models and structure kernels (the panels' sizes and where the parts sit; measured, not copied), ESA's renders and photographs (its finishes; reference only).
• Hayabusa2 (JAXA): our model, from published figures: JAXA's Hayabusa2 Information Fact Sheet v2.31 (2018: the 1 x 1.6 x 1.25 m body, the 6 m paddles, the labelled device views), ISAS's note on its planar high-gain antennas (Hirokawa 2014: 92 cm and 90 cm), the NAIF frames kernel hyb2_v17.tf (its axes and layout), JAXA's paper model and CGs (Akihiro Ikeshita) for proportions and JAXA's photographs of the flight spacecraft (its finishes); JAXA publishes no spacecraft model, and its material is used for reference only (JAXA site policy).
• MMX (JAXA): our model, from published figures: JAXA's MMX pages (5.3 m tall, 9.0 m long, 4.5 m wide; its three modules), ESA's MMX factsheet (its engines and mass), DLR's IDEFIX facts and figures, JAXA's 1:48 printable model of 2024 (where the parts sit; measured, not copied) and JAXA's renders (its finishes); JAXA's material is used for reference only (JAXA site policy).
• Danuri (KPLO; KARI, with NASA): our model, from published figures: the KARI/NASA KPLO Danuri launch press kit (August 2022: 3.18 x 6.3 x 2.67 m, 678 kg, 760 W, its labelled hardware render), JPL Horizons' Danuri header (the 1.82 x 2.14 x 2.29 m bus, from KARI), eoPortal (ShadowCam's 1.18 m barrel, KMAG's 1.2 m boom), Jeon et al. (2023), SpaceOps 2023 #217 (the antennas and the arrays' drives), and KARI's renders (its flight-configuration figure, CC BY 4.0; the Ministry of Science and ICT's, KOGL Type 1), used for reference only; no public 3D model with a clear licence exists.
• Chandrayaan-2 orbiter (ISRO): our model, from published figures: ISRO's GSLV MkIII-M1 / Chandrayaan 2 brochure (3.2 x 5.8 x 2.1 m, 2,379 kg, 1,000 W), eoPortal's Chandrayaan-2 page (the central cylinder, the two solar panels, the dual-gimbal X-band antenna), Bhiravarasu, S. S. et al. (2021), PSJ 2:134 (DFSAR's 1.4 x 1.1 m antenna), and ISRO's photographs, drawings and renders of the flight orbiter (GODL-India), used for reference only; no public 3D model with a clear licence exists.
• ESCAPADE (NASA/UC Berkeley SSL/Rocket Lab), Blue and Gold: our model, one for both twins, from published figures: NASA's ESCAPADE press kit and fact sheet (2025: 4.88 x 1.65 x 1.09 m deployed, 1.20 m wide stowed, the 2 m magnetometer boom, each instrument's place) and the kit's labelled spacecraft render, the 0.6 m high-gain antenna (UC Berkeley SSL, via Wikipedia), NASA's photograph of the two flight spacecraft (their finishes); no public 3D model with a clear licence exists.
• Advanced Composition Explorer (NASA/Johns Hopkins APL): our model, from published figures: the Caltech ACE Science Center's brochure (the two-deck octagon 1.6 m across and 1.0 m high, the dish on the aft deck, four fixed arrays), eoPortal (four 86.4 x 149.9 cm panels, about 8.3 m across), the ACE/MAG project (the magnetometers 4.19 m from the centre), NASA's labelled exploded view and NASA's photograph of the flight spacecraft (its finishes); NASA 3D Resources' Advanced Composition Explorer measured for the layout only (loose faces, no single scale fits it).
• Wind (NASA): our model, from published figures: NASA GSFC's Wind WAVES note (Wilson III; the wires 100 m and 15 m tip to tip, the axial antennas 12 m, the magnetometer boom 45 degrees from them; after Bougeret, J.-L. et al. 1995, Space Sci. Rev. 71:231), NASA GSFC's MFI description (the 12 m Astromast and its two magnetometers; Lepping, R. P. et al. 1995, Space Sci. Rev. 71:207), the drum 2.4 x 1.8 m, NASA's labelled diagram, painting and clean-room photograph of the flight spacecraft (its finishes); NASA 3D Resources' Wind measured, not used (it is Polar's geometry).
• Deep Space Climate Observatory (NOAA/NASA/USAF): our model, from published figures and photographs: NASA (stowed 5.4 x 6 ft, 1,256 lb), NASA/GSFC's labelled drawing (the SMEX-Lite bus, the panels 72 in. tall, EPIC's 30 cm telescope, NISTAR, the Faraday cup, the flat high-gain antenna, the boom), eoPortal (the 1.37 m bus), Loto'aniu, P. T. M. et al. (2022), Space Weather 20:e2022SW003085 (the 4.0 m boom), and NASA/KSC's photographs of the flight spacecraft (KSC-2014-4577, KSC-2015-1240; sizes measured on them, and its finishes); NASA 3D Resources' DSCOVR (Triana) measured, not used (the 1999 concept).
• IMAP (NASA, Princeton University, Johns Hopkins APL): our model, from published figures: NASA's IMAP press kit (2025: 2.4 m across, 0.9 m tall, 797 kg, the 2.5 m magnetometer boom), McComas, D. J. et al. (2025), Space Sci. Rev. 221:100 (the 3.6 m² body-mounted array, twelve 4 N thrusters, the antennas, each instrument's angle to the spin axis and its place round the deck), and NASA's photographs of the flight spacecraft at Astrotech (its blankets and finishes); no public 3D model exists.
• Carruthers Geocorona Observatory (NASA, University of Illinois Urbana-Champaign, UC Berkeley SSL; built by BAE Systems): our model, from published figures: "The Carruthers Mission Concept and Performance" (arXiv:2608.10130: its masses and power, the body-mounted panel, the GeoCoronal Imager on three bipods with its doors and star trackers, COSSMo, the thrusters; its drawings for the layout), NASA/BAE Systems renders, and NASA's photographs of the flight observatory (its sizes, measured against its Lightband); no public 3D model exists.
• Perseverance (Mars 2020): NASA (NASA 3D Resources), Mars 2020 Perseverance Rover; no individual modeller named. In metres as published; the mast raised from the model's own animation; the wheels and suspension split into parts (ours). Sizes checked against NASA's rover components page.
• Curiosity (Mars Science Laboratory): NASA (NASA 3D Resources), Curiosity Rover (MSL) (Clean); no individual modeller named. In metres as published; posed from the model's own animation (mast up, arm stowed). Sizes checked against NASA's Curiosity mission page.
• Ingenuity: NASA (NASA 3D Resources), Ingenuity Mars Helicopter; no individual modeller named. In metres as published; sizes checked against JPL's Ingenuity landing press kit and NASA's photo PIA23882.
• InSight: NASA (NASA 3D Resources), InSight Cruise Lander (arm deployed); no individual modeller named. In metres as published; the instruments set on the ground (ours) from NASA JPL's deployment releases and the InSight launch press kit (arrays east and west, workspace south); the tethers ours.
• Viking lander: NASA (NASA 3D Resources), Viking Lander; no individual modeller named. Scaled 0.2033 to the footpad triangle of 2.21 m (NSSDC, via Wikipedia). Corrected to the published heights (body 46 cm deep, 22 cm ground clearance: NASA Facts, JPL, "Viking Mission to Mars"; camera origins 0.487 m above the deck and about 1.3 m above the ground, 0.822 m apart: NASA PDS, Geometric Calibration of Viking Lander EDR Images); legs rebuilt after NASA CR-112115 (McDonnell Douglas, 1972) with 30.5 cm footpads (NASA Facts); cameras rebuilt 58.4 cm tall (Smithsonian NASM A20050005000); high-gain dish enlarged to 76 cm (NASA Facts). Shapes of the legs, cameras and landing radar, the parts split and named, and the painted finish ours; compared with NASA photos PIA09703, 101-KSC-73PC-630 and 101-KSC-75P-110.
• Spirit and Opportunity (Mars Exploration Rovers): NASA (NASA 3D Resources), Mars Exploration Rover - Spirit and Opportunity; no individual modeller named. In metres as published; posed from the model's own animation; texture colour space fixed (ours). Sizes checked against NASA JPL's MER launch press kit (2003).
• Galileo: NASA/JPL/Solar System Simulator (NASA 3D Resources), Galileo; the deployed high-gain antenna, the atmospheric probe, the plasma-wave dipole, the relay antenna's strut and the science boom's four instruments (EPD, PLS, HIC, DDS) ours, to published sizes and masses (PDS, JPL, UW-Madison SSEC, University of Iowa); the science boom shortened to 3 m and the magnetometer boom attached at its end (NASA PDS Galileo Orbiter host description), the GPHS-RTGs shortened to 1.14 m, the antenna tower to about 1.3 m above the rim; blanket quilting ours; checked against JPL's drawings and the 1983-1989 JPL and KSC photos.
• Juno: NASA (NASA 3D Resources), Juno (A); no individual modeller named. Scaled to the 2.5 m high-gain antenna (NASA PDS Juno host catalogue); the Waves antenna shortened to 4 m tip to tip (NASA/JPL Juno launch press kit, 2011); the cells darkened from NASA's photos of the flight arrays (PIA13928, KSC-2011-3848).
• Pioneer 10: NASA/JPL/Eyes on the Solar System (NASA 3D Resources), Pioneer 10; the plaque's engraving from NASA Ames photograph AC72-1338 (plaque by Carl Sagan and Frank Drake, art by Linda Salzman Sagan); the plaque's mount, the antenna's gap and the RTG truss length set from NASA Ames photograph AC72-2135, TRW's drawing (via Scheffer 2003) and the PDS spacecraft description. TRW Systems for NASA Ames Research Center, "Pioneer F/G: Spacecraft Operational Characteristics", PC-202, May 1971, NASA Technical Reports Server 20190030279, https://ntrs.nasa.gov/citations/20190030279 . Figs. 2-10, 3.1-2 (side view with stations: overall length, low-gain antenna tip, dish rim, horn), 3.1-4 (aft view: louvres, meteoroid-detector panels, thrusters), 3.8-1; section 3.8.1 (SNAP-19: 11.10 in long, 15.68 in across the fins, six fins); 2.1.3 (17 ft four-segment magnetometer boom). Used for the low-gain antenna's length and the RTGs' size. US Government work. R. O. Fimmel, W. Swindell, E. Burgess, "Pioneer Odyssey", NASA SP-349 (1974), NTRS 20190002224, https://ntrs.nasa.gov/citations/20190002224 : 2.9 m length, thruster pairs fore and aft, 1.2 m feed struts, 0.76 m low-gain antenna. NASA photographs used for surface detail (NASA Image and Video Library): ARC-1973-AC73-9019 (Pioneer 10 final assembly at TRW), ARC-1971-AC71-2696 (Pioneer F/G thermal model), KSC-73P-0116 (Pioneer G at Cape Kennedy), with AC72-2135 already credited.
• New Horizons: NASA Visualization Technology Applications and Development (VTAD), New Horizons 3D Model (science.nasa.gov/resource/new-horizons-3d-model/), scaled to APL's 2.1 m high-gain dish and set in NAIF's NH_SPACECRAFT frame (nh_v220.tf); checked against APL's payload diagram and clean-room photo (pluto.jhuapl.edu) and NASA's 3D-print kit (science.nasa.gov/3d-resources/new-horizons/); material tuning ours. G. H. Fountain et al., "The New Horizons Spacecraft", Space Science Reviews 140 (2008) 23-47, arXiv:0709.4288, https://arxiv.org/abs/0709.4288 : propulsion (16 thrusters in 8 sets), thermal louvres on the lower deck, RTG collar fins. Figures saved for reference only. NASA/KSC photographs (NASA Image and Video Library), 2005: KSC-05pd2458 and KSC-05pd2461 (the GPHS-RTG: black finned housing), KSC-05pd2412 (the antenna assembly: collars on the mast and feed), KSC-05pd2414, KSC-05pd2494. "finned RTG housing ... covered with a high emissivity coating"; 42.2 cm fin tip to fin tip, 114 cm long: Space Power 9:1 (1990), https://ssi.org/wp-content/fbs/ssisp_v09n01_1990/54 (for the RTG's colour; sizes as already credited).
• MESSENGER: NASA 3D Resources, MErcury Surface, Space ENvironment, GEochemistry, and Ranging (MESSENGER); solar panels fitted to APL's 1.5 x 1.65 m, split into its moving parts.
• Ulysses: NASA/Brian E. Kumanchik, NASA/Christian A. Lopez (NASA 3D Resources), Ulysses; sizes from ESA's Ulysses spacecraft pages and datasheet (cosmos.esa.int/web/ulysses); axial antenna and wire dipole shortened to ESA's lengths, dish logos painted out (ESA's 1990 photos). Detail pass ours: the bus's gold blankets crinkled, with taped seams every 0.5 m, as ESA's ESTEC photos show them.
• Skylab in the gallery: our model, from published figures: NASA MSFC, MSFC Skylab Mission Report: Saturn Workshop, NASA TM X-64814 (1974; NTRS 19750002066): module sizes (fig. 2-1), the workshop wing (fig. 7-10), the observatory's wings (section 7.3.1), the twin-pole thermal shield (fig. 10-48), the nitrogen tanks (section 4.1.2), the discone antennas (fig. 5-5); NASA EP-107, Skylab: A Guidebook (1973; NTRS 19730025070): overall sizes, the canister, docking ports, optical window, EVA hatch, thrusters, radiator (figs. 70, 71, 81, 86, 98); MSFC Orbital Workshop final report vol. II sect. 2.2.4 (the thruster system: six nozzles, 22 spheres); the Smithsonian's 1:20 Skylab model (NASM A19720973004); shapes beyond those, the finishes (images drawn for the model) and the shield's drape ours, from NASA photos SL4-143-4706, SL4-143-4707, SL3-114-1660, SL3-114-1682, SL3-115-1837, SL2-04-248, SL2-X4-256, SL2-01-111, S73-26773 and MSFC/KSC 7028935, 7023006, 7027629, 9801812. NASA 3D Resources' Skylab was a layout check only. Terms: facts, cited; NASA photos public domain.
• Gemini in the gallery: NASA/Michael D. Carbajal, Gemini (NASA 3D Resources), scaled to the adapter's 10 ft; sizes from Gemini VII's records (This Day in Aviation). Terms: NASA media usage guidelines, credited.
• Friendship 7 in the gallery: NASA/Michael D. Carbajal, Atlas 6 (Friendship 7) (NASA 3D Resources), scaled to the heat shield's 190.5 cm (Smithsonian NASM A19670176000); the finish ours (the shingles' corrugation baked from a procedural pattern), from the Smithsonian's CC0 photographs NASM2016-01952 and NASM2015-05370; height checked against NASA's Results of the First United States Manned Orbital Space Flight (1962), fig. 4-11. Terms: NASA media usage guidelines, credited; Smithsonian photos CC0.
• Saturn V in the gallery: NASA/Michael D. Carbajal, Saturn V (NASA 3D Resources), scaled to 110.6 m; its stages pulled in to their published diameters (Wikipedia's Saturn V and Apollo command and service module); the S-IC's blue band repainted white with four flags, as NASA's Apollo 11 rollout photo (20 May 1969). Detail pass ours: the F-1 and J-2 engines from NASA's fact sheets (F-1: Saturn V News Reference; J-2: MSFC IND B1411E), the skins made round and corrugated with the model's livery kept, stage fittings after MSFC's S-II and S-IVB cutaways (MSFC 67 IND 1200-63A, -65A), the spacecraft and launch escape system from published lengths (Wikipedia's Apollo CSM and Launch escape system) and NASA photo 6900560 (Apollo 11 at the pad); KSC/MSFC photos 6521629, 6862832, 9903405 of the F-1; the S-IC/S-II (aft) interstage split off the S-II as its own part at the model's ring joint 5.45 m above the first separation plane (about 18 ft, 5.5 m, published). Terms: NASA media usage guidelines, credited.
• OSIRIS-REx: NASA/Christopher R. Meaney (NASA 3D Resources, OSIRIS-REx), scaled to the 6.2 m span (NASA OSIRIS-REx fact sheet, 2016); the TAGSAM arm ours, to the fact sheet's 3.35 m, posed as at the touch-and-go; the sampler head scaled to about 30 cm from NASA photo jsc2023e058644.
• NEAR Shoemaker: NASA/Brian E. Kumanchik and Christian A. Lopez (NASA 3D Resources, NEAR Shoemaker), in metres as drawn (the 1.5 m dish of NASA PDS's mission catalogue). Detail pass ours: the aft-deck instruments split out (Cheng et al. and Hartka and Persons, Johns Hopkins APL Technical Digest 19(2), 1998); the solar cells drawn 42 across each panel and the copper string lines as in NASA photo 9701593 (KSC, 1997), the gallium arsenide cells per APL Tech. Dig. 19(2); the blankets' foil crinkled; the 450 N engine's nozzle a smooth bell as in that photo (Mosher and Wiley, APL Tech. Dig. 19(2)); the panels' backs plain dark grey (undocumented).
• Dawn: NASA/Brian E. Kumanchik (NASA 3D Resources, Dawn), in metres as drawn (19.7 m across the arrays, NASA/JPL Dawn at Ceres press kit, 2015).
• Stardust: NASA/Brian E. Kumanchik and Christian A. Lopez (NASA 3D Resources, Stardust), in metres as drawn (5 m long, NASA/JPL Stardust comet flyby press kit, 2004).
• Deep Impact: NASA/Brian E. Kumanchik and Christian A. Lopez (NASA 3D Resources, Deep Impact (EPOXI)), in metres as drawn (the 2.8 m panel of NASA/JPL's Deep Impact launch press kit, 2005); the impactor a separate part.
• Launch Complex 39A in the gallery (as on 16 July 1969, Apollo 11): our model, from published figures: NASA MSFC, Saturn V Flight Manual SA-503, MSFC-MAN-503 (1968; NTRS 19750063889), section VIII and figs. 8-6 to 8-10 (the launcher base, the 380 ft umbilical tower 80 ft north of the vehicle, the arms, hold-down arms and tail service masts, the pad, flame trench and deflector); C. D. Benson and W. B. Faherty, Moonport, NASA SP-4204 (1978); HAER FL-4, Mobile Launcher One (National Park Service, 1984; Library of Congress); the tower's levels and the arms' stations after Apollomaniacs' "Mobile Launcher & Launch Umbilical Tower"; the site's spheres, perimeter and crawlerway placed from OpenStreetMap (© OpenStreetMap contributors, ODbL). The members' layout, the arms' trusses and hinges, the mounts' places, the slopes and the colours ours, checked against NASA photos KSC-66C-6962, 64C-5545, KSC-69PC-443, KSC-69P-446, KSC-69PC-353, 6900560 and the rollout of 20 May 1969. Terms: facts, cited; NASA photos public domain; OpenStreetMap ODbL, credited.
• Sputnik 1 in the gallery: our model, from published figures: NASA NSSDC Master Catalog, Sputnik 1 (1957-001B); RKK Energia's figures as given by A. Zak, RussianSpaceWeb "Sputnik design", and Wikipedia's Sputnik 1 (58 cm sphere, two 2 mm hemispheres, 36 bolts, whips 2.4 and 2.9 m at 35 degrees); antenna bases, brackets, flange and ports ours from the NSSDC replica photo and the US Air Force Museum replica. Terms: facts, cited; NASA photo public domain.
• MSL/Mars 2020 aeroshell: our model, from published figures: Edquist et al., AIAA 2009-4078 (NTRS 20090024230), fig. 1 (shape, backshell cones, features); Edquist et al., AIAA 2009-4075 (NTRS 20090024218); Way et al. (NTRS 20090007730); Mars 2020 press kit; tile layout, finishes and the features' sizes ours, from NASA/JPL photos PIA23590 and PIA24128. Terms: facts, cited; NASA/JPL photos public domain.
• Curiosity's parachute: our model, from published figures: Cruz et al., AIAA 2013-1250 (NTRS 20130012763), table 1 and fig. 3 (as-built disk, gap, band, lines, riser, bridle; fabrics); gore colours ours, from NASA/JPL test photos PIA11981 and PIA11994. Terms: facts, cited; NASA/JPL photos public domain.
• Apollo 11's drogue and main parachutes: our models, from published figures: R. B. West, Apollo Experience Report: Earth Landing System, NASA TN D-7437 (1973; NTRS 19740003586), figs. 2 and A-15 to A-17 (the Increased Capability Block II system: two 16.5 ft conical ribbon drogues, three 83.5 ft ringsail mains; gores, vents, line and riser lengths); the inflated shapes, the cluster's spread, the rings and ribbons, and the colours ours, the colours after NASA photo S71-18753 (Apollo 14's descent; Apollo 11's layout is not published). Terms: facts, cited; NASA photo public domain.
• Perseverance's parachute: our model, from published figures: geometry: Cruz et al., AIAA 2013-1250 (NTRS 20130012763) and the Mars 2020 press kit; the gore code: NASA/JPL-Caltech, "Mars Decoder Ring" (PIA24431), designed by Ian Clark, read gore by gore. Terms: facts, cited; NASA/JPL image public domain.
• MSL/Mars 2020 descent stage and sky crane: our models, from published figures: Way et al., IEEE Aerospace 2007 (NTRS 20090007730: eight MLEs canted 25 degrees, three tanks, 390 kg hydrazine, BUD bridles 7.5 m); Mars 2020 press kit (6.4 m cords); layout, sizes and finishes ours, from NASA/JPL photos PIA24128, PIA11439, PIA11426, PIA11808. Terms: facts, cited; NASA/JPL photos public domain.
• Mariner 4 in the gallery: our model, from published figures: NASA NSSDC Master Catalog, Mariner 4 (1964-077A); JPL Technical Report 32-740, "Mariner Mars 1964 Project Report: Mission and Spacecraft Development", Vol. I (1966, NASA NTRS 19660019455), section III and Fig. 29 (octagon 18 x 54.5 in, panels 71.4 x 35.5 in, vanes 7 sq ft at 35 degrees, dish 46 x 21 in at 38 degrees, mast 88 in, 7 in ground plane, where everything sits); spar and fitting sections, cell layout, dish depth and feed, louvred bays, instrument positions and TV camera size ours, from NASA/JPL photos (P-4732, "Mariner 3 and 4", the Mariner Mars 1964 cutaway). Terms: facts, cited; NASA/JPL photos and report public domain.

• Rosetta (gallery): NASA/JPL/Eyes on the Solar System (NASA 3D Resources, Rosetta), its wings scaled to ESA's 32 m and the high-gain dish to ESA's 2.2 m (ESA, The Rosetta orbiter); Philae NASA's stowed block. Source verified byte-for-byte against NASA's official 3D Resources repository on 23 October 2026. Its collection reuse statement and NASA media guidance support reuse with NASA acknowledgement and no implied endorsement; cleared for the next hosted build.

The Explorer's timeline

The Explorer's timeline marks these and other missions and sky events, each with its own source (NASA NSSDCA, the Apollo Flight and Lunar Surface Journals, NASA and NASA JPL mission pages, ESA, ISRO, the NASA Eclipse Web Site, the JPL Small-Body Database).

Eclipse comparisons

• Earth, 8 April 2024: Eclipse Predictions by Fred Espenak, NASA's GSFC, the central-path table. Its 96 complete rows are retained in assets/eclipse_references/earth_2024.json. This version specifies UT1 and ΔT = 74 seconds. JPL DE440 and Skyfield supply the model's apparent vectors.
• Jupiter, 24 January 2015: Hubble's 06:28 UT image, NASA, ESA and the Hubble Heritage Team, STScI/AURA. The displayed reference is a resized copy of the supplied image. JPL/NAIF DE440, JUP365 and PCK00011 provide the planet, Io, Europa and Callisto geometry.
• Mars, 2 April 2022: Perseverance's Phobos eclipse, NASA/JPL-Caltech/ASU/MSSS/SSI. The comparison uses the original Mastcam-Z frame ZL7_0397_0702168518_882EBY_N0180000ZCAM01517_1100LMJ, NASA/JPL-Caltech/ASU, whose image address and metadata are retained in assets/eclipse_references/mars_2022.json. Rover location comes from the mission's M20 waypoints, RMC 18/0 at the sol-396 endpoint. NAIF's Mars 2020 spacecraft-clock kernel converts the recorded clock; DE440, MAR099S and PCK00011 supply the apparent transit geometry and Phobos ellipsoid.

Software in the app

satellite.js (MIT), Shashwat Kandadai and contributors: SGP4 in the viewer's Live mode.

three-mesh-bvh (MIT), Garrett Johnson: fast raycasts against the shape models (67P and the other irregular bodies) in the viewer.

Lucide (ISC), Lucide Icons and Contributors, including icons from Feather (MIT), Cole Bemis: the viewer's everyday icons (play, zoom, help, layers and the like), redrawn to the viewer's square line. The space glyphs (planets, orbits, the light pulse, the terminator, map projection, the data badges) were drawn for the project.

three.js (MIT): the interactive viewer. Its KTX2 texture loader carries the Basis Universal transcoder, Binomial LLC (Apache-2.0), with ktx-parse and zstddec by Don McCurdy (MIT); the textures are encoded with Basis Universal's basisu (Apache-2.0). The models' geometry is packed with meshoptimizer's gltfpack and decoded by its decoder (MIT), Arseny Kapoulkine.

The film

The twelve moments, 1969 to 2400, and what is on screen in them. The film's end credits are this part, shortened: what each card says is marked here.

Solar Timeline

The past is drawn from the record; the future is our own design.

About the film

A film of the solar system from 1969 to 2400: what was really there on each date, and what might be there next.

Twelve moments: Earthrise from Apollo 11's command module Columbia, the first steps at Tranquility Base, Voyager 1 at Jupiter, a 1985 base on the rim of Shackleton crater, a 1995 landing in Coprates Chasma on Mars, sunrise over the International Space Station with Atlantis docked, Philae's landing on a comet, Cassini's last plunge into Saturn, and four journeys outward to Ceres, Io, Saturn's rings and Neptune.

A non-commercial film, with music written and played for it. Its code is open source under the MIT licence; the film, its music and its own designs are shared under CC BY-NC 4.0.

Accuracy card

The positions of the Sun, Earth, Moon and planets, the phases they show and the sunlight in every shot are computed for that shot's date from NASA JPL's DE440 planetary ephemeris. Planets, moons and the Sun are shown at their true sizes as seen from the camera. Sunlight falls with distance from the Sun as it does in space.

The terrain at Shackleton crater is NASA's laser-altimeter map of the lunar south pole, and the Sun there clears the real skyline.

Philae and Cassini fly their real reconstructed paths, from ESA's and NASA's SPICE kernels. Other spacecraft paths, the camera positions and all hardware after 2017 are designed for the film.

Where the film departs from the record, it does so on purpose. For example: shot 1 runs the orbit at twice real speed; shot 3 places Voyager by hand at its real distance; Ceres uses its mean distance; Earth in shot 10 is enlarged to stay visible and Neptune's rings are drawn brighter than true.

Where everything was

Shackleton's terrain and rocks

Lunar soil and small craters

The ground on Mars

Neptune's moons

Neptune's rings

Saturn's B ring up close

Events and references

The film's look draws on For All Mankind (Sony Pictures Television / Apple TV+) and The Expanse (Alcon Entertainment / Amazon). No designs, footage or names from either are used.

Software

Blender 5.2, Blender Foundation (GPL): every scene modelled, lit and rendered, in Cycles and Eevee.

Flamenco 3.9, Blender Studio (GPL): the render farm.

MPFB 2, Joel Palmius and the MakeHuman Community (GPL): the crew's bodies.

Skyfield, by Brandon Rhodes (MIT), and SpiceyPy, NASA NAIF's SPICE toolkit for Python (MIT): where the Sun, planets, moons and spacecraft were.

rasterio and GDAL (BSD, MIT): the lunar terrain.

NumPy (BSD), Pillow (HPND) and FFmpeg (LGPL/GPL): grading, titles and the cut.

Open source and open data

All the software above is free and open source, under the licences given. The data and models come from free public archives: NASA (JPL, NAIF, the Planetary Data System, 3D Resources), USGS Astrogeology and ESA, mostly in the public domain, with credit owed where marked; Solar System Scope maps are CC BY 4.0; the Smithsonian's scan of Armstrong's spacesuit is used under its non-commercial terms; ESA's archive data (the comet's shape, Gaia's stars) are CC BY-NC 3.0 IGO. No endorsement by NASA, ESA, JAXA, the Smithsonian or any other source is implied. The computers ran Apple's macOS, which is not open source. AI was used to build and manage the render farm and to develop the software; it is a commercial service, not open source.

Maps of the Moon and Earth

Maps of the planets, moons and stars

The Moon maps and the star map are from Internet Archive copies of the NASA SVS files; their sizes match the SVS listings.

Our own designs

• Meridian, an original fictional expedition and rescue cutter for the interactive model review. Geometry, layout, markings and PBR textures are authored in this project. Exterior-detail inspiration: J_the_Ninja's Rocinante modelling breakdown and Master Replicas / Eaglemoss's Razorback model, inspected 22 October 2026. No images or geometry from those references are included in Meridian. This asset does not appear in the released film. The later charcoal, ivory and red livery also draws visual direction from a reference image supplied by Daniel (artist and original source not supplied). Its pixels and geometry are not used in the model or textures. The revision 5 lettering references the worn identity markings in a Syfy Tachi promotional still reproduced by L'Operatore Spaziale, inspected 22 October 2026. Meridian's angular wordmark and Mars meridian/orbit emblem are original drawings; no show insignia, font outlines or reference pixels are included. Daniel's later reference crops informed revision 6's projecting bow probes, heavier lettering and simplified colour fields. The probe structure, tank end fittings and all additional markings were authored here. Revision 7's interior fittings and controls are also original. Its fictional Earth–Mars demonstration uses the existing JPL DE440 endpoint ephemerides with our solar-gravity finite-burn solver. It is not a reconstruction of an Expanse journey; planetary gravity, capture, propulsion engineering and crew tolerance are not modelled.
• Designed for the film: Shackleton Base (1985), the Mars lander (1995), the Ceres polar station (2150), the drive ship (2150-2400) and the Saturn ice hauler (2185). Our own designs, in the spirit of The Expanse and For All Mankind but taken from neither. Their solar arrays are drawn cell by cell to real cell sizes: 2 × 4 cm silicon at Shackleton (the space cell of the 1970s and 1980s) and 4 × 7 cm triple-junction at Ceres (today's space cells); the look was studied from NASA's ISS model (below), and nothing is copied from it.
• The base and the lander are operated in the film by Hawksmoor-Vance Orbital, a fictional British company; its HV mark is our own. "ESA" appears only as a stencilled word; ESA's logo, a trademark, is not used.
• The Union Flag, drawn to its official construction: the national flag, free for anyone to fly or show (College of Arms, Union Flag FAQs, https://www.college-of-arms.gov.uk/resources/union-flag-faqs).
• The European flag (the European emblem), drawn to its official geometric description: owned by the Council of Europe and the EU, and usable by third parties where there is no likelihood of confusion with the EU or the Council of Europe and it is not tied to aims incompatible with theirs (EUIPO, guidelines for third-party use of the European emblem, https://euipo.europa.eu/tunnel-web/secure/webdav/guest/document_library/contentPdfs/about_euipo/resources/the_use_of_the_emblem_en.pdf). Here it is a flag on fictional hardware beside a fictional company's own mark, as the guidelines allow.

Apollo and Voyager

• Neil Armstrong's Apollo 11 spacesuit (A7-L), National Air and Space Museum, Smithsonian Institution; 3D model: Smithsonian Digitization Program Office. © Smithsonian Institution, used for this non-commercial film under the Smithsonian's terms (not CC0). The backpack (PLSS and OPS) is modelled for the film at sizes of ours, estimated from NASA photos (no published dimensions found). In the gallery, the PLSS, OPS, RCU and lunar overshoes are modelled from Smithsonian NASM records A19830164000, A19740611000, A19830157000 and A20010292000, and NASA photos AS11-40-5875, AS11-40-5878, AS11-40-5903 and S69-31048 (public domain).
• Command Module Columbia, Apollo 11 (A19700102000), National Air and Space Museum, Smithsonian Institution; 3D model: Smithsonian Digitization Program Office (CC0). Interior: render-ready model by Autodesk (2015) from the Smithsonian Digitization Program Office scans, prepared for Blender by Max Anderson (2019); CC0.
• Columbia as flown in lunar orbit (gallery): the command module's shape from the Smithsonian scan above (CC0), its forward heat shield, probe, mirror-silver finish and the service module ours, to the Apollo 11 press kit's and NASA's published sizes, placed from NASA photos AS11-37-5445, S69-32370, MSFC 69-MS-G-2288 and AS17-145-22254. Its cabin is the render-ready interior above (Autodesk; Max Anderson; CC0), registered to the scan; also in shot 1's set in the viewer. Its RCS quads' R-4D engines (13.4 in long, 5.6 in exit, canted 10 degrees out, the pod clocked 7) from AAHS, Rocket Propulsion Evolution 9.32 and 9.45 (SM RCS, LM RCS, after NASA's CSM and LM news references); the engines' detail ours.
• Apollo Lunar Module: NASA/Michael D. Carbajal (NASA 3D Resources), scaled up 1.42 times, the ascent stage lowered to the press kit's 6.99 m (Grumman's LM Structures study guide, fig. 2), the landing gear reworked to 9.45 m across the 0.94 m footpads, and rigged to fold as stowed for launch after W. F. Rogers, Apollo Lunar Module Landing Gear (7th Aerospace Mechanisms Symposium, NASA, 1972; NTRS 19730010151), fig. 2; the RCS quads numbered as the Apollo 11 LM RCS report (NASA, Mission Report Supplement 5); the MESA and the plaque ours; the finish, the landed probes, the ladder and the hatch adjusted to Eagle's from NASA photos AS11-40-5927, AS11-40-5902 and AS11-44-6581, and the ladder from NASA OCHMO's Apollo Lunar Lander brief; foil and panel textures ours. The RCS engines rebuilt as R-4Ds (13.5 in, a 5.75 in exit, area ratio 40, the cobalt-alloy extension's eight stiffening rings: AAHS Rocket Propulsion Evolution 9.45) on the source's axes, and a thermal shield on quads 1 and 4 like the source's on 2 and 3.
• Voyager: NASA/Michael D. Carbajal (NASA 3D Resources), Voyager Probe (B).

The ISS and Atlantis

• International Space Station: NASA/JSC Integrated Graphics, Operations, and Analysis Laboratory (IGOAL), via NASA 3D Resources; set back to July 2011 for the film. Raffaello is drawn with the model's Leonardo, the same module shell. Pirs, the two Soyuz and two Progress docked on 16 July 2011 and the Nauka airlock and radiator then stowed on Rassvet: NASA JSC's "ISS complete 2011" models (ISS (C) high res, NASA 3D Resources). Today's station in the Explorer's Live mode: the IGOAL model as published, the current configuration.
• The Cupola's glass, shutters and interior (lining, handrails, robotics workstation, lights): built for the film on the model's Cupola, from ESA's Cupola facts (esa.int, "Cupola": seven windows, the 80 cm top window, the debris, pressure and scratch panes, the shutters) and NASA crew photographs: ISS028-E-016231 (12 July 2011, the shutters open with STS-135 and Expedition 28 crew in the windows), ISS028-E-028792 (August 2011, inside at a window) and ISS036-E-029210 (the Cupola robotics workstation, 2013); the workstation's parts after the Canadian Space Agency's description of the Mobile Servicing System.
• Space Shuttle Atlantis: Orbiter, Space Shuttle, OV-103, Discovery, National Air and Space Museum, Smithsonian Institution; 3D model: Smithsonian Digitization Program Office (CC0). Its sister ship's scan, renamed and set up as Atlantis on STS-135 for the film. Orbiter Docking System and Canadarm: NASA 3D Resources (Space Shuttle Parts); the arm posed as in NASA photo ISS028-E-017801 (18 July 2011). The inspection boom, carrier, pump module and floodlights are ours, from published sizes and NASA photos.
• Positions of Voyager 1 and Jupiter's moons, and the orbit of the ISS: NASA JPL Horizons.
• The crew in Shackleton Base's windows: bodies, clothes and hair from MakeHuman (CC0; Data Collection AB, Joel Palmius, Jonas Hauquier and the MakeHuman Community), built with MPFB 2 and recoloured for the film. Their movement: the Carnegie Mellon University Graphics Lab Motion Capture Database, mocap.cs.cmu.edu (takes 07_01, 13_04, 13_09, 15_06 and 40_10); the database was created with funding from NSF EIA-0196217.

Comet 67P and Philae's path

• Comet 67P's shape: the SHAP4S model from OSIRIS stereophotogrammetry (DLR): Scholten, F., SPG SHAP4S plate model, RO-C-MULTI-5-67P-SHAPE-V1.0 (2016), after Preusker, F. et al. (2015), A&A 583:A33; ESA/Rosetta/MPS for OSIRIS Team MPS/UPD/LAM/IAA/SSO/INTA/UPM/DASP/IDA (ESA, H. Sierks); via the ESA SPICE Service and Planetary Science Archive, CC BY-NC 3.0 IGO.
• Philae's descent (SONC-FD, CNES) and bounce (the 7Dwarfs reconstruction, B. Grieger et al. 2019), the comet's attitude and orbit: ESA SPICE Service, ESA/ESOC Flight Dynamics. The Sun: NASA JPL DE440; the comet's orbit checked against the JPL Small-Body Database.

Philae's landing

• The landing: Biele, J. et al. (2015), Science 349:aaa9816; the landing gear and the touchdown: Witte, L. et al. (2016), Acta Astronautica 125:149; Roll, R. and Witte, L. (2016), Planetary and Space Science 125:12; the feet 2.3 m apart: Boehnhardt, H. et al. (2017), Phil. Trans. R. Soc. A 375:20160248; the ground at Agilkia: Mottola, S. et al. (2015), Science 349:aab0232 (ROLIS); the comet's colour and albedo: Fornasier, S. et al. (2015), A&A 583:A30; the failed hold-down thruster and harpoons: Biele et al. (2015) and Ulamec, S. et al. (2018), JIMIS 4; the touchdown site before and after: ESA Rosetta blog, "OSIRIS spots Philae drifting across the comet" (17 November 2014); Philae's spin: ESA Rosetta blog, "Did Philae graze a crater rim during its first bounce?" (28 November 2014, ROMAP).
• Philae: modelled for the film to the published figures (no ESA, DLR or NASA model exists). Time runs up to 60 times faster after the bounce, shown on screen.

Cassini's Grand Finale

• Cassini's trajectory and attitude on its last orbit: NASA NAIF, Cassini SPICE archive (naif.jpl.nasa.gov/pub/naif/CASSINI/kernels/): reconstructed trajectory 200128RU_SCPSE_17235_17258.bsp, reconstructed attitude 17257_17262ra.bc, frames cas_v43.tf, clock cas00172.tsc; leap seconds naif0012.tls; read with SpiceyPy. NASA, public.
• Saturn's moons where they were (Enceladus and Tethys in frame): NASA JPL satellite ephemeris SAT441 (R. A. Jacobson), via NAIF; Saturn's shape and pole: IAU 2015 (pck00011).
• The final plunge: NASA JPL, Cassini: End of Mission press kit (2017); NASA, Cassini End of Mission Timeline; NASA JPL, Reconstructing Cassini's Plunge into Saturn (2018). SIGNAL LOST is the last telemetry in spacecraft time, at Saturn; Canberra lost the carrier 83 minutes later, at 11:55:46 UTC (E. Lakdawalla, The Planetary Society, 15 September 2017).
• Saturn's rings: the main ring edges from NASA's Saturnian Rings Fact Sheet and Colwell et al. (2009), above; the D ring: Hedman, M. M. et al. (2007), Icarus 188:89; Nicholson, P. D. and Hedman, M. M., in Planetary Ring Systems (2018). Saturn's Bond albedo: Hanel, R. A. et al. (1983), Icarus 53:262.
• Saturn's north polar hexagon and cloud detail (illustrative, after): Cassini ISS images, NASA/JPL-Caltech/Space Science Institute.
• Cassini: NASA/Brian E. Kumanchik (NASA 3D Resources, Cassini-Huygens (A), without Huygens), split into its moving parts. Time runs up to 600 times faster, shown on screen.

Rendered on

A home render farm of two Apple computers running Flamenco: a MacBook Pro (Apple M5 Pro, 48 GB), which also ran the farm, and a Mac mini (Apple M4 Pro, 24 GB), six render workers between them, on their GPUs through Metal.

Geological-unit maps of Jupiter's moons

The Explorer rasterises the USGS GIS unit polygons into streamed maps. Unmapped areas are grey; structural lines and point symbols from the printed maps are not included. Io and Ganymede use labelled app colours. Europa uses the GIS palette and unit names from its pamphlet.

Phobos retains Gaskell's older body coordinates. Deimos retains the historical Thomas frame, with west longitude converted to east and about 400 m source uncertainty over much of longitude 200–355 degrees. Their placement in the current rotation is nominal; the maps are not precision-registered to the shapes.

Argonaut illustrative CAD (local review)

ESA and Thales Alenia Space: Argonaut programme and CAD download, Argonaut_TASI_Newlander_01.stl, retrieved 23 October 2026. ESA generated the passenger; the descent element is inspired by Thales Alenia Space's concept. The page offers the model for illustration and asks for these credits. Converted-mesh redistribution is not explicitly granted; public_ok: false keeps the model out of public builds pending clarification under ESA's website terms. Local preparation reduces 1,770,870 triangles to approximately 240,000, centres the terrain base at zero height and adds a neutral grey review material. STL supplies neither materials nor articulated parts. Units are unspecified: source scale is retained provisionally, without equating this concept's 13.21 × 13.18 × 9.37 scene envelope (terrain base, astronaut and mast included) to the newer published flight design. Mission target: 2030, Ariane 64, lunar south polar region; no precise landing site, date or flight path is invented.

Nancy Grace Roman Space Telescope model

NASA Goddard / Conceptual Image Lab: textured deployed illustration, RST_Model_V004.glb, from the Roman commissioning page. Used under NASA media guidelines, with no endorsement implied. Adapted by centring, uniformly scaling to NASA’s approximate 12.7 m overall length, naming components and retaining source textures and geometry. This is illustrative geometry, not flight CAD: its scaled mirror is 2.12 m versus the real 2.4 m primary. Original source SHA-256: a873fc95945ed0992c552c2a97fc6c5cd0bf7f5519df0040d5a299b03bae67ce. The web copy omits two unusable normal/occlusion map references on the black insulation: the exported mesh has no UV coordinates. Its colour, metalness, roughness and geometry are retained.

• MMS · Magnetospheric Multiscale: NASA / Christopher R. Meaney, NASA 3D Resources, from its official NASA GitHub mirror. NASA media usage guidelines. Source textures and geometry retained; NASA’s textured illustration, shared by MMS 1–4. Uniformly scaled by two to an approximately 3.5 m instrumented bus. The radial wire booms are shortened in the source model; this is not a dimensionally complete flight model or measured attitude.

• THEMIS / ARTEMIS probe: NASA / JPL / Eyes on the Solar System, NASA 3D Resources, from its official NASA GitHub mirror. NASA media usage guidelines. Source textures and geometry retained; NASA’s educational illustration, shared by the THEMIS/ARTEMIS probe family. Original geometry, textures and metre-scale units retained. Radial wires are shortened in this source: the actual deployed EFI spans are 49.6 m and 40.4 m. This is not flight CAD or a measured attitude.

• XMM-Newton: our original geometry and baked material images, from ESA Science Programme achievements, pp.163–165 and ESA’s mirror technical description. Published envelope, tube and panel dimensions and mirror-shell counts retained; smaller fittings, blanket pattern, shell curvature and radiator geometry illustrative. ESA photographs used as references only; no ESA source mesh or image redistributed.

• Dragonfly preparation (unassembled, local only): Gary Allen Backus / Backus Creative Imaging, NASA 3D Resources printable kit, published 16 October 2026. NASA supplies a 1:23.5 educational model reconstructed from publicly available imagery, not flight CAD. Original 3MF and guides retained locally; our preparation converts units and simplifies copies of its 14 print parts. Source archive and previews are not included in public builds. Assembly, materials and flight-layout checks remain outstanding.

• SOLAR-1 draft (held, local only): our original geometry using NOAA’s mission illustrations, NASA KSC-20250725-PH-BAE01_0016 hardware photography and NCEI’s instrument description as references. NCEI gives the two magnetometer distances as approximately 5 m and 6.6 m. Bus/panel dimensions and instrument layout remain illustrative estimates; the draft is excluded from library auto-discovery. No reference photography or third-party geometry is embedded in it.

Friendship 7 and DSN observatory paths

Friendship 7: NASA Manned Spacecraft Center, *First United States Manned Three-Pass Orbital Mission*, NASA TM X-563-1 (1964), Table XXI, actual post-posigrade state (public report copy). Positions are our numerical integration with central Earth gravity and J2, not recovered telemetry. Atmospheric drag, higher harmonics, solar/lunar gravity and control impulses are omitted. Starts after separation rockets and stops before retrofire; no launch or entry trajectory is invented. Timing: NASA’s 1962 flight results, Table 6-II. The fireflies milestone uses David G. Simpson’s NASA case study; its educational orbital elements are not used. Capsule attitude is illustrative.

MMS 1–4 and XMM-Newton: NASA/JPL Horizons Earth-centred ICRF vectors, queried at 30-second intervals for the dated tours. Four independent MMS trajectories share the family model. These are dated paths, distinct from the live DSN feed; model attitude is not flight telemetry.

Library orbit coverage

SOHO, ACE, Wind, DSCOVR, IMAP, Carruthers, Roman, TESS, Chandra, Mars Express, Hope, Danuri, Chandrayaan-2, ARTEMIS-P1/P2 (THEMIS-B/C), Kepler and Spitzer: NASA/JPL Horizons geometric ICRF vectors. Target identifiers, dates and sample intervals are in tools/flybys.py (LIBRARY_ORBITS); raw responses are cached under data/horizons_flybys. These are dated trajectory solutions, potentially combining reconstructed and predicted segments, not live telemetry. The L1 paths are shown relative to Earth; Roman's span includes its transfer towards L2. Kepler and Spitzer are shown relative to the Sun during their operating missions. Attitudes are illustrative.

Drive ship at Ceres and ice hauler at Saturn: our original fictional demonstrations in tools/fictional_orbits.py, two circular coasting orbits each. Gravitational parameters come from JPL's DE440 GM kernel; solar lighting comes from Horizons. Point-mass gravity only, with no burns, docking, stationkeeping, ring-particle avoidance or third-body perturbations. They are fictional examples, not engineering feasibility claims or agency plans.

Satellite catalogue enrichment

Readable catalogue owner/source and launch-site names use CelesTrak’s SATCAT owner/source dictionary and launch-site dictionary, retrieved 24 October 2026. These describe catalogue codes, not necessarily the current operating company. The shared snapshot is assets/satellite-info.json; unknown codes remain visible. Mission descriptions are original short summaries of the NASA mission pages linked on each entry (Terra, Aqua, Aura, Landsat 8/9, Swift and IXPE). They describe mission purpose, not current operational status.

Terra and Aqua Earth-observation models (25 October 2026)

• Terra: NASA / Christian A. Lopez, source GLB. Uniform scale 0.0005 makes the source bus span 6.81 m, checked against NASA’s 6.8 m length.
• Aqua (A): NASA / Kevin K. Lane, source model. Downloaded the actual aqua-(a)/Aqua%20(A).glb, not the page’s mislabelled Aquarius link. Uniform scaling to NASA’s 16.7 m deployed span retains source proportions (4.77 × 16.7 × 8.83 m versus published 4.8 × 16.7 × 8 m).
• NASA media usage guidelines; credit NASA and the named artists, no endorsement implied. Source hashes and download URLs are recorded in assets/models.json; source geometry/materials retained, static poses, no claim of measured attitude. Catalogue IDs 25994 / 27424 from CelesTrak Earth Resources.
• Landsat 8 source inspected but not integrated: NASA / Christopher R. Meaney’s downloadable GLB has an exploded assembly. Reassemble and validate it before replacing its generic satellite.

Jezero landing-area detail (Explorer, 2026-10-25)

NASA/JPL/USGS HiRISE Mars 2020 terrain-relative-navigation orthomosaic, Fergason, R. L., Galuszka, D. M., Hare, T. M., Mayer, D. P., and Redding, B. L. (2020), DOI 10.5066/P9QJDP48. USGS data release marks it CC0 1.0. The JPL surface-operations derivative is JEZ_hirise_soc_007_orthoMosaic_25cm_Ortho_blend120.tif, from the USGS S3 distribution.

tools/mars_deep.py reads a 1.2 km window centred on 18.4446°N, 77.4509°E, transforms its orthographic Mars-sphere coordinates to the Explorer's planetocentric map, and builds sparse tiles down to about 0.6 m/pixel. The source is 0.25 m/pixel. The imagery stays greyscale, with the original image's lighting baked in. Edges blend into the globe. This first patch does not displace terrain and is not a claim of wheel-level registration or coverage of the entire traverse.

The landing-area relief now uses the matching 1 m HiRISE DEM: JEZ_hirise_soc_006_DTM_MOLAtopography_DeltaGeoid_1m_Eqc_latTs0_lon0_blend40.tif from the same USGS S3 distribution. Fergason et al. (2020), 10.5066/P9REJ9JN, USGS data release, CC0. tools/mars_deep.py --fetch-dem reads a bounded 1.3 km source window. A 257-square grid samples this at approximately 5 m spacing, with centimetre quantisation and a smooth taper to zero at the patch boundary. Height at the landing point is subtracted: **relative relief anchored to the reference ellipsoid**, not an absolute areoid-to-ellipsoid conversion. The source datum and subtracted height remain in surface.json. Ground mesh, rover placement, rover tilt and camera clearance use that same grid. Model wheels and suspension are rigid; this is not a reconstruction of measured wheel contacts or attitude.

The local regional extension (tools/mars_regions.py) adds a second window centred on the recorded 2 July 2021 traverse position (18.4360984 N, 77.4517056 E), using those same image/DEM releases and CC0 terms. Both DEM windows use the original landing reference height, not independent local zeroes. Fine samples share a 5 m cylindrical lattice (east-west ground spacing varies with latitude), with centimetre quantisation. The union coverage fades at its outer edge; overlapping regions do not independently fade towards zero. A bounded coarse field backs the fine fields, with smooth residual blending at their edges. The current two-region set is loaded before scene creation; mesh, rover/path placement and camera clearance share that fixed snapshot. This extends early-traverse coverage only, not the whole route. Original image lighting remains baked in; wheels and suspension remain rigid.

Apollo LRV engineering greybox (review only, 2026-10-25)

tools/assets/apollo_lrv_greybox.py is original geometry built from published envelope dimensions in NASA NSSDCA, The Apollo Lunar Roving Vehicle: frame length 3.1 m, wheelbase 2.3 m, tyre diameter 0.818 m, tyre width 0.23 m, and loaded ground clearance 0.36 m. The greybox's track width, folding axes, seat and equipment shapes, antenna, suspension struts and guards are provisional layout, not dimensioned flight-hardware reconstructions. No third-party model, texture or Don McMillan imagery was incorporated. It is not in the Library, mission paths or film. The Boeing LRV Operations Handbook is the next engineering reference for correcting those provisional assemblies.

Shuttle launch-part review imports (2026-10-25)

NASA/Johnson Space Center, Space Shuttle Parts: External tank.glb and Solid Rocket Booster.glb, used under the same NASA Image and Media Usage Guidelines as the project's other Shuttle parts; source credit retained, no implied endorsement. shuttle_launch_parts_review.py preserves source geometry and textures, provisionally converts inches to metres, turns the long axis upright and centres the base for inspection. Scale and mission configuration are not validated; these are review assets, not in the Library or film. Source checksums and cross-section reports are emitted with the review builds. See reports/Shuttle launch asset audit.md for discrepancies and required work before use.

Perseverance science stops (Explorer, 2026-10-25)

assets/drive_stories.json holds original summaries and links to NASA's first-road-trip map, Roubion assessment and sample catalogue. The July 2 view is an illustrative stop on the recorded early campaign route. August 6 and September 6 are date-level sampling/sealing context, not exact operation timestamps. Navigation chooses noon UTC on each date and interpolates the existing MMGIS traverse; it does not simulate drilling. The on-screen notes state this and link directly to the sources. Roubion is described as an unsuccessful rock core with an atmospheric sample, not a successful rock sample. No source images or audio were added by this feature. High-resolution terrain remains restricted to the existing two Jezero windows.

Historical mission audio candidates (review only, 2026-10-25)

NASA Historical Sounds, under its linked Images and Media Usage Guidelines: STS-135 Countdown to Launch, STS-135 Launch Commentary, and Mercury 6 Zero G. Exact download URLs and SHA-256 values are in assets/archive_audio.json. NASA is the credited source; no endorsement is implied. These are edited archive excerpts, not continuous mission loops. The unmodified MP3s are cached for review; temporary mono PCM is used only for local machine transcription. Draft transcripts are unverified and are not published or used as captions. No clips from this candidate set have been added to viewer playback yet.

Juice's August 2024 lunar–Earth gravity-assist paths use geometric ICRF vectors from NASA/JPL Horizons (spacecraft −28), sampled at 2 seconds near closest approach and 60 seconds farther out. The existing Juice model is reused. Encounter validation uses ESA's flown cruise report, ops v470, 14 April 2026, not the earlier trajectory-assessment planning table. Context: ESA's 21 August 2024 flyby report. The viewer's radial height above its ellipsoid differs slightly from ESA's reported altitude; the measured differences are 0.4 km at the Moon and 0.2 km at Earth, with closest-approach times within one second. Attitude and Director cameras are illustrative; the two encounter windows omit the intervening coast.

Europa Clipper's Mars encounter preparation uses NASA/JPL Horizons (-159), geometric ICRF positions relative to Mars, for an eight-hour window on 1 March 2025. Samples are 2 s near encounter and 60 s farther out. The 25 February NASA/JPL overview gives rounded pre-flyby comparison figures (17:57 UTC, 884 km), not exact reconstructed states. Horizons yields 17:56:54.0 UTC and 885.5 km radial height over the viewer's Mars ellipsoid; it is not forced to match those publicity figures. The displayed 11.16 km/s is Mars-relative, whereas NASA's 24.5 km/s approach figure is heliocentric. The cached Horizons object summary identifies JPL navigation solution cruise006-final-OD055 across this encounter. The post-flyby infrared report and radar report support the instrument-test context. No exact instrument-operation timeline or measured spacecraft attitude is claimed; the existing model is reused.

The Apollo canopy deployment greybox (tools/assets/apollo_canopy_deployment_review.py) reuses the main-canopy dimensions documented by the existing Apollo parachute builder from NASA TN D-7437, Figure A-15. Its 15% and 40% mouth fractions, restricted canopy bulge and profile interpolation are authored illustrations, not measured reefing ratios or a cloth simulation. Authored meridians preserve constructed length. An adaptive 11-pose bake limits sampled intermediate cloth and suspension shortening to 1 cm (seven checks per interval; independent regression at 31 weights per interval), rather than claiming exact preservation. It is an isolated single-canopy review asset. Its 68 suspension lines join the authored skirt at the published line length in each review pose. The 5 mm line radius and single convergence point are illustrative. Ringsail slots, risers, deployment bag, pilot and a mission deployment clock remain absent.

Landsat 8 original-source review: NASA / Christopher R. Meaney, LDCM-2014-composite.lwo from the official NASA-3D-Resources repository at commit 0ec49456c824c0d42049bdd62940a11a0a617731, path 3D Models/Landsat 8 (LDCM)/LDCM-2014-composite.lwo. SHA-256 d728dce6963b0e6a126576d5786386c4806b550b214e2183f8b638872dec71a7. The NASA model page credits the author and links the official repository. Our pinned, single-layer LWO2 review importer reads original polygons and flat surface colours, exchanges Y/Z axes and reverses winding, applies the same provisional scale as the published GLB node, and records Blender mesh validation. Procedural LightWave materials are not reproduced; the model remains an unassembled review asset.

Curiosity's day-level Director science stops use NASA's First Sample Drilling for John Klein on 8 February 2013, and Farewell to Murray Buttes, Image 2 for the 8 September 2016 view and the interpretation of cross-bedded sandstone as deposits from migrating dunes. Stop coordinates remain interpolated from the existing recorded traverse. Neither local outcrop geometry nor drilling operation timing is reconstructed by these stops. No source photographs are reproduced as new textures.

Curiosity — John Klein ground imagery

A bounded 300 m patch uses Calef, F. J. III and Parker, T. (2016), Gale merged orthophoto mosaic, 25 cm, NASA/JPL-Caltech/USGS/University of Arizona. USGS-produced data are public domain; attribution retained. The source is MSL_Gale_Orthophoto_Mosaic_25cm_v3.tif. tools/gale_deep.py reads a bounded window from the strip-organised source, checks its registration, coverage and checksum, then builds sparse map tiles. The centre is interpolated from Curiosity's recorded route at the day-level John Klein science stop. Greyscale lighting is baked into the photograph; there is no local height field or reconstructed drilling/outcrop geometry. This does not provide detailed imagery for the entire traverse.

New Glenn launch-vehicle greybox (review only)

Original geometry in tools/assets/new_glenn_review.py, informed by Blue Origin's New Glenn overview and NG-2 mission account, checked 25 October 2026. The nominal 98 m envelope, 7 m fairing, seven BE-4s, two BE-3Us, four control fins and six landing legs constrain this review of the seven-engine vehicle, not the separately advertised 9x4 variant. The body's 7 m diameter, individual stage lengths, fairing curve, engine geometry and arrangement, strake outlines, hinge locations and deployment travel are provisional modelling choices. No logos or downloaded source mesh are used. The Blue Origin photograph newglenn_giant-among-rockets.jpg was consulted privately for visual reference; it is not redistributed in the asset. This is not an NG-2 flight reconstruction, a verified launch stack or a timed path.

The LRV review wheels now use original sparse wire geometry, simplified chevrons and an internal bump-stop ring. NASA NSSDCA's same LRV reference gives 0.083 cm steel-strand diameter and 64.8 cm bump-stop diameter; these dimensions are used directly. The 64 strands per diagonal family, twenty chevrons, chevron sweep, rim sections and spokes are representation choices, not verified flight counts or construction. The wheel shape is rigid, with no load deflection.

Gale colour and elevation preparation (staged only)

The same USGS Gale orthophoto catalogue now offers the July 2025 colour product by Timothy Parker and Fred J. Calef III, MSL_Gale_HiRISE-LRGB_78quads_sharp_cog.tif. Its colour strips were registered to the greyscale mosaic and sharpened by the publishers. Bounded crops for both Curiosity science stops are staged separately; original lighting and published processing remain baked into the pixels. Credit NASA/JPL-Caltech, USGS and University of Arizona; catalogue use constraints: none, credit authors.

tools/gale_relief.py stages matching windows from USGS Gale merged DEM v3, Parker and Calef, 2016, MSL_Gale_DEM_Mosaic_1m_v3.tif. This merges multiple instruments tied to MOLA. One-metre grid spacing is not one-metre registration accuracy. Source elevations require an explicit local datum and seam/alignment checks before use on the viewer ellipsoid. No terrain activation is implied.

The isolated Gale terrain review models use those same registered colour and DEM windows. tools/assets/gale_relief_review.py samples a 129-square mesh at about 2.34 m spacing, keeps source registration, and subtracts each crop's centre elevation for a local review frame. Original image lighting remains baked; review lighting adds shading. The two local datums are not interchangeable and these meshes are not yet active Explorer terrain.

The animated single-canopy review now includes TN D-7437's 78-inch steel, 15.5-inch fabric and 42-inch connector axial lengths. Its six steel cables use the published 0.28-inch diameter. Cable spacing, connector envelope and fabric cross-section remain provisional geometry; no deployment mechanics are claimed.

Apollo 15 surface preparation (tools/apollo15_surface.py, staged only): NASA/GSFC/Arizona State University LROC NAC Apollo 15 DTM, NAC_DTM_APOLLO15.TIF (2 m posts), registered NAC_DTM_APOLLO15_M111578606_50CM.TIF imagery (0.5 m), and NAC_DTM_APOLLO15_CONF.IMG, directly from the LROC PDS archive LRO-L-LROC-5-RDR-V1.0/LROLRC_2001. Public-domain PDS products under the LROC terms; credit NASA/GSFC/ASU. The product readme reports 1.06 m relative precision and 1.74 m LOLA RMS; these are not claims of wheel-scale registration. Original image lighting remains baked. Crops use the published LROC Apollo 15 landing coordinate, 26.13239° N, 3.63330° E, from its temporal traverse. This replaces the rounded site definition, about 23 m away; the visible LM now lies at the crop centre. Pixel size does not imply placement accuracy. Confidence values in the fetched v1.9 product exceed the published readme's category lookup; the crop is preserved as uninterpreted, not a validity mask. No terrain or imagery activation is claimed.

Landsat 8 model preparation now uses that pinned original LightWave source through tools/assets/real_landsat8.py. Comparison with NASA's published model preview confirms the source layout; the earlier separated-hardware concern was not grounds for reassembly. No components were repositioned relative to one another. The source panel measures 9.774 m, within 0.3% of the nominal 32 ft (9.7536 m) in NASA SVS 11490. The source spacecraft axis is rotated from +X to Blender +Z. Degenerate triangles are removed; flat source colours remain, without recreating LightWave procedural materials. Credit NASA / Christopher R. Meaney. The model represents Landsat 8 only, in a static illustrative pose.

Chandrayaan-3 coverage preparation (tools/chandrayaan3_coverage.py): NASA/JPL Horizons targets -158 (lander) and -169 (propulsion module), Moon-centred ICRF vectors, archived ISRO/JPLNAV solutions. Raw queries and responses remain in data/horizons_flybys/ch3_review_*. Availability includes predicted continuation and does not establish flown landing accuracy. The diagnostic uses the viewer's IAU Moon rotation, not a high-precision frame fit. ISRO's mission details and LVM3-M4 brochure provide hardware specifications and reference views for future original models; no ISRO artwork or mesh is redistributed by this preparation.

Pragyan's original greybox (tools/assets/pragyan_greybox.py) uses ISRO's Chandrayaan-3 specifications and labelled rover views and the LVM3-M4 brochure, page 6 for the published 917 × 750 × 397 mm envelope and mechanism identification. Geometry and plain review materials are authored here; no reference artwork, logos or downloaded mesh are embedded. Proportions, wheel/tread geometry, rocker-bogie links and the folded-envelope interpretation remain provisional. The two panel poses are review states, not measured deployment timing.

Pragyan's wheelbase and staged surface landmarks: Kannan V. Iyer, Medha S. Alurkar, Ajay Kumar Prashar, K. Suresh and Amitabh (ISRO Space Applications Centre), Path tracing of the Chandrayaan-3 Rover from OHRC images, Current Science 128(6), 558–561 (2025), DOI 10.18520/cs/v128/i6/558-561. Figures 4/5 give 680 mm wheel-centre spacing front to rear and 750 mm across outer wheel faces. Table 2 supplies eight inferred turn locations, stored as factual coordinates in assets/pragyan-stops.json. No article figures, text passages or OHRC imagery are redistributed. The table does not supply event times or the complete traverse; no continuous measured drive is claimed.

Gale neighbourhood preparation (tools/gale_neighbourhood.py) uses the same USGS Gale merged DEM v3 as the registered relief reviews. Coarse fields sample that 1 m source at 18.75 m projected spacing around John Klein and Murray Buttes; detailed centres retain the reviewed crop. Each field is independently anchored at its science-stop centre, with a feather to the viewer reference ellipsoid at the outer edge. They are relative-relief preparations, not absolute areoid heights or an obstacle envelope. No new imagery licence or provider is introduced.

Pragyan APXS review mechanism: the Physical Research Laboratory's APXS instrument description and hardware photographs identify the six source apertures, central silicon drift detector, fixed four-metal calibration target and stowed/deployed configuration. PRL gives about 5–6 cm source/detector clearance during surface observation. Our original review geometry uses a 55 mm endpoint and a named deployment pivot; the housing size, pivot location, target shapes and motion between endpoints are proportional illustrations, not engineering dimensions or mission telemetry. No PRL photographs, graphics or logos are embedded or redistributed.

Pragyan APXS operation times: PRL's Observations with APXS and linked operation summary provide observation/mobility IDs, UTC start/end, exposure and availability. assets/pragyan-operations.json retains factual values and a source-image hash; no source table artwork is redistributed. tools/pragyan_operations.py joins matching mobility IDs to the separately credited Iyer et al. turn coordinates. Operation intervals include calibration overhead and cannot establish exact instrument deployment times, traversing speeds or the intervening route.

Apollo 15's source registration review (tools/assets/apollo15_relief_review.py) uses the PDS crops credited above through the same bounded mesh/UV builder as Gale. The 300 m patch is centred at the published LM coordinate and offsets source elevations by the centre sample; it does not treat PDS elevations as viewer ellipsoid heights. Its 2 m source posts and 0.5 m image pixels do not establish wheel-scale positioning. Confidence-map encoding remains unresolved; no quality-validated terrain or rover route is asserted. Source image shadows remain in the texture and review lighting adds shading.

SLIM orbital preparation (tools/slim_coverage.py): NASA/JPL Horizons target −240 metadata and vectors, post-launch JAXA trajectories, Moon-centred ICRF. The audit excludes documented defective intervals and does not use predicted descent as a flown landing. An additional discontinuity at 17 January 2024 00:00 is retained as a rejected review window. Coarse chord-speed screening is not a navigation-accuracy claim. JAXA's SLIM technology page links a GLB and the JAXA site policy; no source GLB is redistributed or optimised by this preparation.

SLIM orbital review (tools/slim_orbit_review.py) compares cached JPL Horizons positions with JAXA's 25 December 2023 lunar orbit insertion result and 15 January 2024 descent preparation update. JST milestone times are converted to UTC by subtracting nine hours. Plots are our own rendering of numerical ephemerides; no JAXA illustration or mesh is embedded. Heights use the canonical lunar mean sphere in tools/bodies.py, not local terrain. These are internal source reviews, not active film shots or app paths; source seams are retained and flagged, not smoothed away.

SLIM's original layout greybox (tools/assets/slim_greybox.py) uses the JAXA project dimensions (2.4 m height, 2.7 m width, 1.7 m depth) and the ISAS technology/component descriptions (two main engines, twelve attitude thrusters, five crushable contact absorbers, structural combined propellant tank). The labelled external-layout image was inspected privately as a reference; neither it nor the restricted downloadable GLB is embedded or redistributed. All mesh geometry and plain materials are original. Internal dimensions, equipment placement, cell-sheet outline and strut attachments are provisional proportions, not a measurement claim. The model has no flight-attitude animation, landing/crush simulation or LEV release operation. It remains outside the Library and shots pending review.

The selectable Gale terrain fields use the already-credited USGS merged 1 m Gale DEM (tools/gale_neighbourhood.py), with 1.2 km coarse neighbourhoods and 300 m fine patches at the two existing Curiosity science stops. Each field is relative to its own stop-centred datum and fades to the reference surface at its coarse boundary. Source elevations from different stops are never added or merged. The 2.34 m fine sample spacing is not improved source accuracy; small boulders, overhangs and wheel contact are not reconstructed. The existing colour imagery retains baked lighting. tools/gale_fields.py installs only validated, already-staged grids and preserves Jezero's default field.

The isolated Apollo canopy deployment review now carries five lower-ring openings and 68 continuous radial seam tapes through all authored morph poses. Ring count, 4% inter-ring openings and 24 mm tape width reuse illustrative choices from our existing canopy model; the new 8 mm drawing offset avoids coplanar surfaces. None of these is a measurement from TN D-7437. The final skirt band remains closed. No flown reefing ratios, inflation dynamics or mission timing are asserted.

IMERG precipitation imagery

NASA PPS/GPM's custom daily rain/snow-rate imagery, served by NASA GIBS (IMERG_Precipitation_Rate). Units mm/hr; this is not a rainfall accumulation, forecast or a specific scientific daily product. The 0.1-degree source is coarser than its 2 km display tile grid. Near-polar estimates can be missing; transparent pixels include no-data and rates below 0.1 mm/hr. NASA GIBS capabilities and its GPM_Precipitation_Rate colour map checked 25 October 2026. Sources: https://gpm.nasa.gov/data/imerg and https://raw.githubusercontent.com/nasa-gibs/worldview/main/config/default/common/config/metadata/layers/multi-mission/imerg/IMERG_Precipitation_Rate.md Open NASA Earth science data under the existing EOSDIS/GIBS terms. Rain and snow legends sample NASA's logarithmic colour map; the upper bin is ≥53.0884 mm/hr. Known daily gaps from the checked capabilities remain unavailable.

The Apollo LRV steering review (apollo_lrv_motion_review.py) uses original review geometry and ideal flat-ground four-wheel rolling. NASA's NSSDCA rover description supports opposite front/rear steering. Our 6 m chassis-centre turn radius and 2 m travel are illustrative inputs, not NASA's wall-to-wall turning limit or mission telemetry. Steering linkage, soil slip, suspension motion and failure states are not reconstructed. NASA's Apollo 15 rover account records initial front-steering trouble: this dual-steering review must not be applied indiscriminately to Apollo 15's first EVA. Four stepped review poses are deliberately not presented as a continuous historical manoeuvre.

The LRV review now also supports ideal single-axle steering. The turning centre moves to the fixed axle; all four wheels still roll, and the non-steering pair stays straight. These are mechanism demonstrations only, not recovered Apollo 15 commands, steering-link geometry or a timed failure/recovery sequence. The rear-only still sheet uses a 6 m lateral turning-centre offset; its chassis-centre circular radius is therefore larger than 6 m.

Atlantis parking review (25 October 2026)

The unpublished open-bay parking review reuses the Atlantis model and sources credited under “The ISS and Atlantis”. It returns the authored RMS joints to their existing rest pose and places the OBSS at the existing cradle coordinates. This endpoint is illustrative: no historical transfer motion, latch operation, clearance certification or launch readiness is claimed. Ku-band remains deployed and the doors remain open. No source geometry or flight asset changes.

Shuttle support-roll mechanism review

The unpublished atlantis_mpm_motion_review.py uses the 31-degree support rotation described in NASA-CR-193473, printed pages 11–12 (PDF pages 16–17), Figures 1–2: https://ntrs.nasa.gov/api/citations/19940012243/downloads/19940012243.pdf. The source drawing was inspected directly. Existing approximate pedestal geometry supplies a provisional pivot; isolated baseline/half/full/repeated poses illustrate the separate support motion, not verified stowage or mission timing. The original Atlantis source credits still apply.

Apollo linked deployment cluster review

apollo_cluster_deployment_review.py instances the existing single-canopy review three times, retaining its NASA TN D-7437 dimensions and documented illustrative profile/slot/line limitations. The cluster spread is shared with real_apollo_main_parachutes.py through cluster_spread; this is an authored clearance layout, not measured flight geometry. Inflation is simultaneous, with no mission-time, pilot/bag extraction or aerodynamic claim.

Apollo pilot-chute greybox

Original review geometry from NASA TN D-7437, Figure A-19 (printed A-13, PDF page 34), inspected directly: nominal diameter, 12 gores, suspension-line and combined/steel riser dimensions. Source: https://ntrs.nasa.gov/api/citations/19740003586/downloads/19740003586.pdf. The inflated diameter ratio, spherical profile, vent, eight rings, gaps, seam and cable thicknesses are authored proxies; the mesh does not claim the published constructed porosity. No source imagery or third-party mesh is redistributed. Bag, bridle, mortar and event timing are not yet modelled.

Half-hour IMERG precipitation imagery

The separate 30-minute layer uses NASA GIBS IMERG_Precipitation_Rate_30min, its published 30-minute time ranges and the same GPM precipitation-rate colour map. The 25 October 2026 capabilities advertise coverage from 1 January 1998, with half-hour gaps at 13 September 2025 20:00 UTC, 22 January 2026 00:30 UTC and 10 February 2026 00:30 UTC. Historical selections use their own half-hour; recent selections are capped at least four hours behind now and a missing slot is not silently replaced. NASA describes Early Run latency as about four hours; GIBS delivery may lag further. This is a half-hour mean rate (mm/hr), not accumulated rainfall, live weather or a forecast. The same NASA open-data terms and GIBS acknowledgement apply. https://gpm.nasa.gov/data/imerg https://gibs.earthdata.nasa.gov/wmts/epsg4326/best/1.0.0/WMTSCapabilities.xml

Apollo 7–10 and 12–17 launch-date milestones and short mission summaries use NASA's individual mission overview pages, accessed 25 October 2026: https://www.nasa.gov/mission/apollo-7/ through https://www.nasa.gov/mission/apollo-17/ . Added launch milestones use day precision, not an asserted midnight launch time. Apollo 11 retains its existing, separately sourced launch time. These events do not establish mission-specific flight paths, hardware variants or landing reconstructions.

SLIM lunar orbit, 10 January 2024

The orbit-only Explorer path uses JPL Horizons target -240 relative to the Moon, 10–11 January 2024 UTC, sampled every two minutes. Its time window is owned by slim_coverage.WINDOWS; the screened interval avoids Horizons' flagged defects and predicted descent. JAXA's 25 December orbit insertion release provides an independent approximate 600 × 4,000 km, 6.4-hour reference: https://www.jaxa.jp/press/2023/12/20231225-1_j.html . Trajectory: https://ssd.jpl.nasa.gov/horizons/ ; target restrictions: https://ssd.jpl.nasa.gov/api/horizons.api?format=text&COMMAND=%27-240%27&MAKE_EPHEM=%27NO%27 . No JAXA model is redistributed, no greybox is promoted, and no measured attitude, landing reconstruction or manoeuvre timing is implied. Height is relative to the app's lunar reference surface, not local terrain.

Chandrayaan-3 orbit-only paths use JPL Horizons targets -158 (Vikram, 17 August 2023 14:00 UTC to 18 August 00:00) and -169 (propulsion module, 23–24 August 2023), relative to the Moon in ICRF. Samples are 60 and 120 seconds respectively. These bounded windows are shared with tools/chandrayaan3_coverage.py. ISRO’s mission chronology confirms separation on 17 August, followed by deboosting. The views do not reconstruct separation or touchdown, and show no unreviewed spacecraft model. The descent source becomes nearly stationary before the reported landing time and remains excluded. Orbit data: https://ssd.jpl.nasa.gov/horizons/.

NOAA aurora preparation: tools/aurora_forecast.py validates the latest OVATION feed at https://services.swpc.noaa.gov/json/ovation_aurora_latest.json. Product explanation: https://www.spaceweather.gov/products/aurora-30-minute-forecast (accessed 25 October 2026). NOAA describes a model forecast of auroral location and intensity; lead time depends on upstream solar-wind travel, and the Kp fallback may have no lead time. Feed Observation Time is not an issue time. The retained integer field is not a cloud forecast or guaranteed visibility. The browser layer uses the same source schema. It shows model values of 5 or more (lower values remain in the downloaded grid), with a green/yellow/red key. It is a forecast painted on the globe, not simulated auroral curtains or an observation. Historical dates and stale data are hidden. Globe and Map views use the same north-first grid, sampling at integer-coordinate pixel centres. Updates are requested at most every five minutes while Earth is shown in Live mode. Source grid: 65,160 bytes; RGBA texture: 260,640 bytes.

Apollo development milestones: NASA mission overviews for Apollo 1, Apollo 4, Apollo 5 and Apollo 6, checked 25 October 2026. Apollo 1 is a ground-test event with day precision, not a flight. Apollo 4 and 5 use the stated UT launch times. Apollo 6 uses minute precision because the overview gives both 01 and 01.5 seconds. Its shortened mission description is also checked against NASA’s “55 Years Ago: The Flight of Apollo 6” (https://www.nasa.gov/history/55-years-ago-the-flight-of-apollo-6/). These are original summaries and timeline milestones, not new trajectories.

Opportunity geological Director stops use original summaries of NASA/JPL’s Endurance rim report (6 May 2004), NASA’s Eagle–Victoria traverse caption and Endeavour arrival report. They reuse the credited PDS traverse and MER model; no source imagery is copied. Endurance uses the report date, not an asserted arrival time. Victoria uses 27 September 2006 from the traverse caption; the contemporary JPL release’s image caption instead says 26 September. These remain day-level navigation cues, not precise operations or detailed reconstructions of crater walls. Endeavour’s arrival date is 9 August 2011 (report published 10 August).

Victoria crater source-window review uses the HiRISE stereo DTM DTEEC_021747_1780_022380_1780_U01 and matched orthorectified RED image ESP_022380_1780_RED_C_01_ORTHO, from the HiRISE product page. Credit: NASA/JPL-Caltech/UArizona. HiRISE states its images are public domain (usage policy). tools/victoria_source.py reads a bounded 1.3 km window centred on the existing Opportunity Victoria science-stop position. Source scale is 1.0118528506825 m/pixel. DTM heights are relative to the Mars 2000 equipotential datum, not the viewer ellipsoid. Cropped imagery and DTM retain distinct affine origins; they must not be aligned by array index. Reserved/saturated imagery pixels remain masked. Review contrast is stretched independently; no terrain or imagery has yet been activated.

Opportunity registration audit (tools/opportunity_registration.py) compares our existing lander-tied conversion with the Transverse Mercator basemap CRS published in the Analyst’s Notebook map documentation. That page also documents separate HiRISE DEM/ortho rectification. The legacy CSV label does not specify its projection: the alternative conversion is a hypothesis for review, not an applied correction or confirmed source frame.

Tianwen-2 launch milestone: CNSA, mission update of 6 July 2026 (https://www.cnsa.gov.cn/n6758823/n6758838/c10760422/content.html), which states launch from Xichang on 29 May 2025. Date only; no launch time or spacecraft ephemeris inferred. Blue Moon Mark 1 planned milestone: NASA launch schedule, updated 21 October 2026 (https://www.nasa.gov/event/clps-flight-blue-origins-blue-moon-mark-1/), checked 25 October 2026. Year-only 2027 plan, not a flown event or dated landing trajectory. These entries use original summaries; no source photographs or models are redistributed.

Opportunity registration source review: PDS Geosciences Analyst's Notebook, traverse resources, MERB_traverse_5130.zip. The archive's DBF provides geographic LON/LAT in a Mars 2000 sphere frame (radius 3,396,190 m), with six decimal places. All 21,104 records match the legacy CSV identity and raw/corrected position fields. Archive SHA-256: 6d099b39ccba6e46ee057e34d8d205215a0c09f393948e57ce21fa8860d9ee92. The label describes this derived traverse as not archived in PDS; it is provided by the PDS node from science-team corrections, not continuous measured telemetry. No terrain or active route change is made by this source audit.

Opportunity route update: the converter now uses the verified archive's LON/LAT fields directly rather than inferring a local XY frame. The route's starting position follows that same source; the independent historical lander catalogue is unchanged. Sol-to-UTC timing remains an interpolation at 14:00 LMST, not measured drive times. Raw HiRISE terrain is not activated by this correction.

Endurance terrain source review: PDS Geosciences Analyst's Notebook, traverse and crater terrain products, ASCII grid, accompanying header. Source hash 3350e148ba2209f46f64c3dd2a69203dbb2cb805eaa23a240ffd5a6a6c6183f3. 558 × 588 samples; centres span X 667.72–834.79 m and Y −219.43–−43.23 m. The header explicitly gives projection Unknown, datum NONE; these coordinates are not assumed to be the traverse XY frame. Source elevations are −8.4938 to 14.150756 m in an unspecified datum. Review only; no application activation, texture reuse or new imagery coverage claim. Verify original authorship/reuse terms and local-to-geographic registration before an application terrain asset.

Apollo 13 rescue milestones use NASA SP-4029's Apollo 13 timeline, GMT column at second precision: accident 14 April 1970 03:07:53, transearth injection ignition 15 April 02:40:39, splashdown 17 April 18:07:41. The mission report's rounded GET sequence agrees when added to its 11 April 19:13:00 range-zero epoch. Do not add the liftoff fractional-second offset again. Narrative context: NASA mission overview. These are timeline entries, not a spacecraft ephemeris, an Aquarius/Odyssey model variant or synchronized archive audio. Earth/Moon actions open context.

Spirit geological Director stops use published image dates, not precise science-operation times: Bonneville in Color (12 March 2004), Spirit atop Husband Hill (panorama 23 August 2005, article 1 September), and Spirit Hits a Home Run (Barnhill image 7 February 2006, publication 17 February). Original short summaries link to NASA/JPL sources; photographs are not copied. Existing interpolated traverse and MER model are reused. No detailed crater/hill terrain, new registration or reconstructed instrument operation is implied.

Hubble aperture-door operation reuses the existing credited model and its Hubble_ApertureDoor hinge from tools/assets/real_hubble.py. The builder's open pose is −75° about local X; the closed pose is −180°, so the operation adds a relative −105° rotation without moving the hinge. Five-second easing is an illustrative mechanism demonstration, not measured actuator timing or a scheduled Hubble event. No geometry, source imagery or model licence changed.

Juice pointing operations reuse the credited model's Juice_MGA_APM and Juice_SolarArray_PY / Juice_SolarArray_MY pivots from tools/assets/real_juice.py. Local Z antenna rotation (±20°) and local Y wing rotation (±15°) are conservative illustrative demonstration ranges, not verified flight limits. Five-second transitions are illustrative timing; this is neither deployment reconstruction nor flight telemetry. The fixed high-gain antenna and deployed instrument booms are unchanged. No new mesh, source image or licence is introduced.

Curiosity ridge Director stops use NASA's Vera Rubin Ridge panorama caption (image acquisition 22 June 2017, not its September publication date) and Gediz Vallis route caption (study-area arrival 14 August 2023). Original summaries link to these sources; no panorama is downloaded. Day-level navigation reuses the recorded traverse, existing rover and terrain fallback. Detailed ridge outcrops, panorama camera orientations and instrument operations are not reconstructed.

MRO solar-wing operations reuse the credited NASA model and existing MRO_SolarArray_PX / MRO_SolarArray_MX pivot origins in tools/assets/real_mro.py. NASA's 2005 launch press kit, page 34 describes the gimbaled solar arrays. Our local-X ±25° rotation and five-second easing are illustrative, not operational limits, flight telemetry, deployment or a full two-axis Sun-tracking solution. The antenna and radar remain fixed. No new geometry or source imagery is added.

Europa Clipper wing operation reuses the existing NASA model and its local-X Clipper_SolarArray_PX / _MX pivots, including wing-mounted REASON antennas. The NASA spacecraft press kit describes the wings and radar arrangement; the mission design paper describes array tracking and holding during observations. Our ±15° motion and five-second easing are illustrative, not flight limits, deployment, telemetry or a Sun-tracking solution. HGA and magnetometer boom stay fixed. No new mesh or source imagery is added.

Juno/Galileo arrival Director cues reuse the existing trajectory and paired orbit-insertion markers in tools/flybys.py; no event time is introduced or copied. Juno retains the distinction between the path's event times and the Earth-reception times written in its markers. The camera sequence is authored, not measured spacecraft attitude or engine telemetry. Arrivals with a missing, reversed or out-of-range endpoint do not receive a burn-window sequence.

MESSENGER arrival guidance uses rounded spacecraft-event times from APL's post-insertion announcement: main-thruster firing at 20:45 EDT on 17 March 2011 for approximately 15 minutes, represented as about 00:45–01:00 UTC on 18 March. Both markers and the burn caption say approximate. The 21:10 EDT reception of shutdown confirmation is 01:10 UTC on Earth, not the spacecraft shutdown time. This rounded window is not the planned 00:45:00–00:59:43 interval in the pre-insertion press-kit figure, nor a claim to reproduce the final attitude-stabilisation tweak. Existing Horizons positions are unchanged; only camera guidance and event notes change.

Calculated spacecraft communications pointing (0.0.9 correction)

Explorer uses geometric Earth and Sun directions at the displayed date, from its existing ephemerides. These are nominal communications poses, not recorded attitude kernels, contact schedules, light-time/aberration-corrected antenna solutions, or reconstructions of science and thrust manoeuvres. Roll is illustrative. Unsupported missions retain their existing orientation rules; authored landing, drive, staged and thrust attitudes take precedence.

• Voyager 1/2: fixed dish/body Earth-pointing, following NASA's spacecraft description. The prepared NASA model's Blender +Y boresight becomes glTF -Z (tools/assets/real_voyager.py). The scan-platform pointing is not reconstructed.
• New Horizons: fixed dish/body Earth-pointing communications pose, following NASA's Jupiter encounter press kit, Communications. Its prepared NAIF-frame model has Blender +Y/glTF -Z boresight (tools/assets/real_new_horizons.py). This does not assert Earth-pointing during observations or reproduce its spin state.
• STEREO-A: Sun-facing body and single-axis Earth-pointing HGA, following the NASA mission booklet, also documented with the local model references in tools/assets/real_stereo.py. Model Blender +Z faces the Sun; the HGA pivots about X and its deployed boresight is +Y. No gimbal stops or thermal off-pointing are reconstructed.
• BepiColombo: nominal radiator-away-from-Sun cruise body with independently steered MPO dish, following Mangano et al., BepiColombo Science Investigations During Cruise and Flybys and ESA's antenna pointing account. The existing model's Blender -Y faces the Sun, and MPO_HGA points along +X (tools/assets/real_bepicolombo.py). Steering is ideal geometric aiming; actual gimbal limits, spacecraft blockage, SEP attitudes and communication windows are not modelled. The existing cruise-stack model limitations remain.

No model geometry or textures were replaced; their existing source/licence credits continue to apply. Pointing rules and glTF axes live in the shared viewer/spacecraft_pointing.js, used by both path and Solar System views.

Pioneer and Ulysses calculated communications poses (0.0.10)

Pioneer 10/11's nominal Earth-pointing dish follows NASA SP-446, Pioneer: First to Jupiter, Saturn, and Beyond, attitude-control description. Ulysses follows the archived mission description, The Ulysses Mission, section 5.1. These are calculated geometries using the existing ephemerides and credited models, not recorded attitudes. The model scripts define dish boresight as Blender +Z, exported as glTF +Y. Spin phase, pointing drift, manoeuvres and post-contact attitude are not reconstructed. No new model or imagery is used.

Cassini's Explorer final-approach path now also uses the existing NASA/JPL reconstructed attitude kernel 17257_17262ra.bc, with cas_v43.tf and cas00172.tsc. Quaternion samples map the credited spacecraft model into J2000; the viewer then applies the body's rotation. Covered intervals are interpolated separately. The 66-second coverage gap and the brief interval after the kernel ends hold the last inertial pose and are labelled as gaps. The film's existing attitude reconstruction is unchanged.

Juno's 30 December 2023 Io flyby uses NASA/JPL NAIF's reconstructed juno_sc_rec_231229_231230_v02.bc, JNO_SCLKSCET.00212.tsc and juno_v12.tf. Five-second quaternion samples preserve the recorded spin and pointing in the existing six-hour path. These are sampled flight attitudes, not commanded Earth-pointing; no new imagery or geometry is introduced. The inputs can be fetched explicitly with python tools/attitudes.py --fetch; builds do not fetch.

OSIRIS-REx's Bennu TAG path uses NASA/JPL's reconstructed orx_sc_red_201020_v02.bc, ORX_SCLKSCET.00118.tsc and the existing orx_v14.tf. The CK header identifies reconstructed ACS data as its input. The body attitude is sampled every five seconds, retaining the path's one-second samples during contact and back-away. Model articulation remains the existing illustrative operation; the recorded data controls the spacecraft body, not the arm joints.

Explorer's Coprates replay reuses our existing fictional Mars-descent shot 5: tools/shots.py, its authored GLB anchor animation and shots/05_mars.json camera frames. It is the film's lander/camera path, not a historical mission or measured trajectory. The existing illustrative canyon is retained. Film exhaust, dust and child-part animation are not reproduced by this transform-only player.

Shackleton's Explorer replay uses our existing shot 4 camera samples and fixed base transform. Its 70 m rim move and pan follow the authored film camera; the fictional base and previously credited NASA LOLA terrain are reused. Crew, moving subparts and the film's practical-light effects are not replayed.

Philae's Explorer film replay reuses shot 11's existing GLB anchor animation and camera samples from shots/11_philae.json, with the local comet frame from shots/11_philae_path.json. The existing Philae trajectory, model and 67P source credits above apply unchanged. This is the film reconstruction, including its accelerated post-touchdown time and illustrative attitude/contact motion, not a new measured attitude product. Dust, added boulders and leg flex are not replayed. No new external asset or licence is introduced.

Voyager's Explorer film replay reuses shot 3's existing 03.glb model transform and all 720 camera samples in shots/03_voyager.json. The model and Horizons position retain their existing credits. This film shot holds the craft at its encounter position while the camera arcs and moves towards the dish; it is not a new continuous encounter trajectory or recorded attitude product. Flight samples subtract the shot body's centre displacement before conversion into the fixed Explorer body frame. No external asset or licence was added.

Cassini's Explorer film replay reuses the existing shot 12 GLB anchor and 768 camera samples from shots/12_cassini.json. Its trajectory and NASA reconstructed attitude retain the film's existing source credits; no new trajectory or attitude was invented. The authored 600×-to-real-time pacing is retained. Explorer holds the shot's reference date for the surrounding scene; atmospheric effects and loss-of-signal graphics are not replayed. Reference rotation axes retain extra precision for the long relative plunge.

New Horizons Jupiter encounter attitude (0.0.10 isolated candidate)

NASA PDS NH-J/P/SS-SPICE-6-V1.0 (DOI 10.17189/1520109), reconstructed engineering-telemetry attitude: NAIF archive, ck/merged_nhpc_2007_v006.bc, sclk/new_horizons_3381.tsc, fk/nh_v220.tf. Prepared NASA New Horizons model already uses that spacecraft frame; glTF -Z (dish boresight) maps to spacecraft +Y. Candidate displays only three hours either side of Jupiter closest approach, with five-second seed samples and adaptive refinement; outside this sampled window it explicitly holds the nearest boundary pose. This is a display budget restriction, not a gap in the source CK. No historical communications mode or Earth-pointing assumption is substituted. Source files remain local build inputs; the viewer receives sampled quaternions.

New Horizons Arrokoth encounter attitude

The same NASA PDS New Horizons SPICE archive supplies merged_nhpc_2019_v029.bc, labelled ACTUAL, KEM1, spacecraft frame -98000. Uses the pinned new_horizons_3381.tsc clock and nh_v220.tf frame. Full coverage of the existing four-hour Arrokoth path, 1 January 2019 03:33:22–07:33:22 UTC; five-second seed sampling with adaptive refinement. No communications-pointing assumption or trajectory change. The CK is 22,542,336 bytes, SHA-256 57604592021b1e7f24e3be516bedd6f7cc93c3a9726f95326011a50bf39e7363.

New Horizons Pluto encounter attitude

NASA PDS NH-J/P/SS-SPICE-6-V1.0 merged_nhpc_2015_v039.bc, labelled ACTUAL / PLUTO ENCOUNTER, spacecraft frame -98000, with pinned new_horizons_3381.tsc and nh_v220.tf. Only the four hours centred on the existing Pluto closest-approach epoch are sampled (14 July 2015, 09:48:29–13:48:29 UTC); outside, the boundary pose is explicitly held rather than presented as measured pointing. Five-second seed sampling with adaptive refinement, no trajectory or model changes. CK size 105,793,536 bytes, SHA-256 91c4a08ce75ed12510731c51b4a9410f6c1f30462bad09aad428b498c951627c.

The separate Charon encounter path uses this same reconstructed 2015 CK and spacecraft frame, sampled independently about its existing closest approach: 14 July 2015, 10:02:22–14:02:22 UTC. Its 4,081 samples fit the unchanged 5,000 limit. Outside this interval the boundary pose is explicitly held. No trajectory or model changes. Independent source probing measured .011050 degrees maximum interpolation error across 26,641 probes; this is not a continuous error bound or a statement of telemetry accuracy.

Dawn Ceres recorded attitude

The six-hour interval around the existing 6 March 2015 capture marker uses NASA/JPL NAIF Dawn reconstructed spacecraft CK, DAWN_203_SCLKSCET.00091.tsc and dawn_v16.tf from the same archive. Pinned input hashes and sizes are in tools/attitudes.py. Operational telemetry quaternions are sampled at five-second seeds with adaptive refinement; the 4,321 samples fit the existing 5,000-sample budget. Outside the interval the boundary pose is held and labelled. The NASA model retains nominal dish alignment (about .0243 degrees from the FK boresight); arrays and ion engines are static. No thrust activity, measured gimbal motion or absolute pointing accuracy is implied. The model-frame audit is in reports/dawn-dish-frame-audit.json.

Juice measured attitude during the 2024 lunar–Earth flyby

ESA SPICE Service's measured spacecraft CK, juice_step_241112_v01.tsc and juice_v46.tf supply housekeeping telemetry orientation. The sampler explicitly uses JUICE_SPACECRAFT_MEAS, avoiding the planned-data fallback of the spacecraft switch frame. Separate four-hour intervals use 721 samples each (20-second seeds, adaptive refinement available). Independent 7,200-point comparisons per interval give maximum source differences of .000501 and .001232 degrees; these are not absolute attitude accuracy. Outside coverage the boundary pose is held and labelled. Our model's solar arrays and medium-gain antenna remain static and nominally aligned; their measured joint motion is not reproduced. No ESA reference CAD mesh is shipped.

Europa Clipper's Mars encounter body attitude uses NASA/JPL's reconstructed EHA quaternion kernel, clock europaclipper_00023.tsc and frame kernel v16. The source comments describe engineering-telemetry time tags typically delayed by 0.25 seconds, sometimes up to 0.5 seconds. Numerical interpolation agreement does not remove that timing uncertainty. The existing NASA VTAD mesh uses a nominal identity glTF-to-spacecraft body mapping, verified against signed body landmarks; this is not optical/RF boresight calibration. Arrays remain in the model's static pose, not the available recorded array-joint motion. Pinned input hashes and source/geometry audits are retained in reports/clipper-*.

Lucy encounter body attitudes use NASA/PDS's released weekly operational SPICE products: lcy_sc_rel_231030_231106_v02.bc, lcy_sc_rel_250414_250421_v02.bc, lucy_v13.tf and lucy_sclkscet_00038.tsc. Source comments explicitly allow interpolation across intervals up to 301 seconds. Agreement with these products is not continuous measured telemetry or absolute pointing accuracy. The existing project model uses its nominal body/dish alignment; its instrument pointing platform and solar arrays remain static. No recorded joint motion is claimed. Input hashes, nominal model landmarks and independent source comparisons are recorded in reports/lucy-attitude-coverage.json.

Historical Atlantic cyclone positions use NOAA National Hurricane Center HURDAT2, the pinned 1851–2025 release dated 23 September 2026. These are revisable post-storm analyses, not live observations or forecasts. The layer shows source positions from the preceding 72 hours without interpolated centres or warning areas. Non-synoptic event times and missing wind/pressure/radius values are preserved. The region excludes other ocean basins; missing prepared records do not establish absence of storms. See reports/cyclone-atlantic-source-audit.json for source hash and NHC format reference. US Government data; NOAA attribution, no endorsement.

Dated two-day flood classifications (candidate)

NASA LANCE MODIS Aqua+Terra MCDWD_L3_NRT, doi:10.5067/MODIS/MCDWD_L3_NRT.061, and VIIRS JPSS1+JPSS2 VCDWD_L3_NRT, doi:10.5067/VIIRS/VCDWD_L3_NRT.002, via NASA GIBS/ESDIS. Product overview. The candidate uses MODIS through 23 June 2025 and VIIRS from 24 June 2025, within the pinned advertised service intervals only. It shows the selected day and preceding day as a composite, not the selected hour or water depth. Clouds, shadows and reference-water errors limit interpretation. Transparent pixels do not establish no flooding. Recurring-flood code2 was unpopulated in RevE; its presence in the legend does not establish later implementation. No endorsement, warning or forecast claim. Source hashes and regional review are recorded in reports/flood-source-audit.json and reports/flood-regional-review.json.

Daily SO₂ column imagery

Optional dated Suomi NPP OMPS upper-troposphere/stratosphere imagery uses NASA GIBS NMSO2-PCA-L2-NRT; product DOI, NASA archive. NASA/NOAA imagery provided by NASA GIBS / ESDIS under the EOSDIS open data policy; no endorsement. The 17 km centre of mass is a retrieval assumption, not measured plume altitude. Daily column in DU, not ash or ground concentration or a warning. 25 km sensor resolution differs from the 2 km display grid. Transparent means missing data; coverage gaps and the pinned end date are retained. NOAA-20 is not substituted.

• Bounded Starlink family labels: NASA FY2023 Aeronautics and Space Report of the President, printed221–222, identifies launch batches2023-056/067 as V2 Mini. Exact identities joined to CelesTrak SATCAT; hashes and exclusions in reports/starlink-active-batch-review.json. Family identification does not establish serial geometry or attitude. Unmatched identities retain the generation-unverified label.

BepiColombo's bounded planned Mercury capture path (20–22 November 2026) uses NASA/JPL Horizons geometric vectors from Mercury, ICRF, provided by ESA's bc_mpo_fcp_00236_20181020_20270411_v01 trajectory. The source is predicted after 28 September 2026, not recorded capture telemetry. Planning status follows ESA arrival updates. Frame conversion uses the existing Mercury pole/prime-meridian model; MTM is omitted without simulated separation motion. This window does not bridge the existing cruise endpoint. Source hashes and numerical limits are recorded in reports/bepicolombo-capture-schema-review.json.

Illustrative Mars and Titan ground skies (Explorer)

Texture-free gradients use the selected date's existing Sun direction. Qualitative references: Curiosity sunset, PIA19401 (NASA/JPL-Caltech/MSSS/Texas A&M Univ.) and Huygens surface, PIA07232 (NASA/JPL/ESA/University of Arizona). Images remain reference-only; no image pixels are shipped or fitted. Mars's localized blue dusk and Titan's diffuse warm haze are illustrative, not measured weather, radiance or calibrated RGB. Smooth fades by 60 km (Mars) and 200 km (Titan) are visual limits, not atmospheric boundaries. Earth's shader is unchanged; Map, Full light and airless bodies do not use these skies.

Home discovery feature previews (2026-10-28)

• Deep Space Network: PIA23214, Goldstone antennas photographed 11 February 2020, courtesy NASA/JPL-Caltech. Resized WebP; JPL image-use policy. A photograph, not a live feed.
• Earth winds: Solar Timeline Explorer screenshot with the NOAA GFS 10 m wind layer enabled, run 28 October 2026 12 UTC +6 h. NOAA NCEP data via NOAA Open Data Dissemination (public domain); NASA Blue Marble composite base. A static feature preview, not current weather. Existing layer credits and forecast-time display remain in Explorer.