A preprint turns faint nighttime passes into measurements
A research team has published quantitative optical measurements of large satellites illuminated after they entered Earth's shadow. In a preprint posted September 7, the authors report 147 nighttime detections of the International Space Station and one observed pass of the Chinese Space Station using a 0.6-meter telescope in Sydney. Sunlight did not directly illuminate the spacecraft during those observations. Instead, the available light came from the Moon and from moonlight reflected by Earth toward the satellites, a contribution the paper calls lunar-Earthshine.
The measurements were collected from June 8 through June 12, 2025, around a nearly full moon. September 7 is the date the manuscript became public, not the observing date or a new spacecraft event. The paper has been submitted to The Journal of the Astronautical Sciences, according to its arXiv record, but the posted version is a preprint and is not identified there as accepted or peer reviewed. Its value is therefore a documented experiment and model comparison, not an operational declaration or a settled claim that nobody had ever seen a moonlit satellite before.
A fast telescope followed two stations through shadow
The campaign used Macquarie University Observatory's 24-inch PlaneWave telescope, an aperture of about 0.6 meters, with a camera and direct-drive mount capable of following fast-moving objects. The researchers attempted every ISS pass that rose above 40 degrees from their site during the campaign and reported a detection each time. They actively tracked the spacecraft rather than leaving the telescope fixed on the stars, using current orbital elements to guide the mount. Individual exposures typically lasted between one and ten seconds, with settings adjusted for target and sky brightness.
For the ISS, the median measured nighttime brightness was V=12.02, with a stated photometric calibration uncertainty of 0.17 magnitude and a one-magnitude spread among observations. The nighttime sample was about 13.1 magnitudes fainter than the separate daylight sample used for comparison. Because astronomical magnitudes run backward, the larger number means a much fainter target. The Chinese Space Station was measured during one 99-percent-moon pass at a median magnitude of 13.42. The authors did not build a detailed model for it because the campaign provided only that limited observation and public structural information was insufficient.
Moonlight reaches a spacecraft by two routes
Direct moonlight is sunlight reflected once from the lunar surface and then from a satellite toward the telescope. Lunar-Earthshine adds another reflection: moonlight illuminates Earth, scatters from the surface or atmosphere toward the spacecraft, and then returns to the ground observer. Which path matters depends on the positions of the observer, satellite and Moon, as well as the spacecraft's orientation and reflective materials. A panel facing Earth may receive little direct moonlight while still being illuminated from below by the moonlit planet.
The team extended the open-source lumos-sat brightness model to include both lunar contributions. Across the 147 ISS nighttime measurements, the difference between observed and modeled brightness averaged 0.03 magnitude with a scatter of 0.80 magnitude. A separate Ansys Systems Tool Kit simulation represented the ISS with more components. For one June 9 pass, that simulation found lunar-Earthshine dominated illumination of modeled surfaces facing Earth and raised their radiance far above the starlight-only case. These are model results tied to an assumed geometry and surface treatment, not direct measurements of light landing on each ISS component.
The experiment supports longer tracking windows
Optical satellite tracking commonly benefits from twilight geometry, when an observer is in darkness while a low-orbit spacecraft remains sunlit. Once the satellite moves into Earth's shadow, operators may lose that strong illumination even though the object is still above the horizon. Detecting reflected moonlight provides another observing regime. For large targets near full moon, the Sydney results show that a modest research telescope can obtain useful optical signals at times that would otherwise be dismissed as too dark.
That extra window has practical value. More opportunities to measure a spacecraft's position can help maintain an orbit estimate, reduce gaps between observations and provide additional chances to inspect brightness changes. The paper demonstrates detectability and photometry, however, not a complete operational surveillance service. It does not publish a new catalog-maintenance accuracy, custody rate or automated identification score. Reliable tracking also depends on weather, pass geometry, telescope scheduling, orbital predictions and the ability to distinguish targets in a growing population.
The authors' near-continuous 24-hour result comes from combining actual observations with a brightness model for full-moon conditions. They did not operate the telescope without interruption for an entire day. The campaign contained only the ISS passes available above Sydney, and the paper explicitly identifies broader temporal coverage as future work. Its modeled monthly analysis suggests suitable large satellites could exceed the telescope's one-sigma detection threshold at local midnight for as many as 11 nights in a representative 30-day month. That is a conditional forecast, not a measured year-round duty cycle.
The calibration sets limits on precision
Several instrument details keep the reported brightness from being treated as universal. The camera lacked a standard photometric filter, so the team calibrated it with 321 reference stars and limited their color range to reduce mismatch. The same calibration parameters were then applied across the observing nights. The authors also warn that the dome moved in discrete steps and may occasionally have blocked part of the telescope aperture while following a fast target. They consequently treat the measured magnitudes conservatively because partial vignetting would make a spacecraft appear fainter.
Near zenith, the field rotator could not always keep pace. Some portions of passes were lost when the ISS left the camera field or became elongated, and bright stars sometimes contaminated frames. The researchers visually inspected every image before photometric use. Those precautions improve the reported sample, but they do not remove the missing portions or turn the instrument into a uniform survey system. A filtered camera, improved dome motion and observations at more lunar phases are needed to establish how accurately another site or telescope can reproduce the result.
A measured signal now anchors future testing
The three-panel lead figure captures the scale of the challenge without functioning as a controlled brightness comparison. It shows the ISS in daylight, twilight and moonlit darkness, but the panels come from different observations with different illumination geometry and spacecraft attitudes. The right-hand moonlit view is visibly faint and noisy, yet the tracked station remains detectable. Quantitative comparisons belong to the calibrated photometry across many frames, not to a visual judgment of those three panels alone.
The constructive result is narrow and useful: reflected lunar light produced measurable signals from two large occupied stations during a five-night, near-full-moon campaign. A model incorporating moonlight and lunar-Earthshine reproduced the ISS brightness distribution with substantial scatter and can now be tested against observations from other sites, lunar phases and spacecraft. Whether this becomes dependable round-the-clock tracking will depend on those replications and on operational metrics the preprint does not provide. The work advances optical tracking by replacing a plausible illumination path with measured data, while leaving universal coverage and performance as questions for the next campaign.