Rubin's test data turns up one of the faintest dwarf galaxies on record

The search started from an assumption about age. The team ran computer models over the test catalog, looking for groups of stars whose brightness and colors suggested they had evolved together, and co-author Aashay Pai said the models treated those stars as roughly 13 billion years old ^[9]. That assumption is the filter. With the age fixed, the model could tell a co-evolved group apart from the unrelated stars and galaxies sitting in front of and behind the search area ^[9].

An eight-sigma threshold answers one narrow question: whether these stars are a chance alignment of unrelated objects ^[10]. The archival check came from a different telescope and a different camera, which argues against an artifact in Rubin's own pipeline ^[11]. Neither test settles what kind of object this is, and Pai said follow-up observations will be required to confirm its identity ^[16]. Joshua Simon, an astronomer at the Observatories of the Carnegie Institution for Science who was not involved in the work, called the discovery robust ^[16].

Pai estimated the mass of Aquarius IV at about 1,500 suns, against 1.5 trillion for the Milky Way on NASA's figure, a factor of a billion ^[14][2]. Its absolute magnitude is -1.9, or 14 million times dimmer than the Milky Way ^[15]. The analyses put it about 359,000 light-years from the galactic center, roughly 3.6 times the length of the Milky Way itself ^[13][3]. Pai described it as "one of the faintest and most compact UFDGs at a large distance" ^[12].

The dark-matter part of this story comes from the class, not from a measurement of this object. Simon said of ultrafaint dwarfs that "these galaxies are composed almost entirely of dark matter: 99 to 99.9% of their mass is dark, and only 0.1 to 1% is made up of stars" ^[5]. They hold at most a few hundred thousand stars, where the Milky Way has around 100 billion ^[4]. The study describes Aquarius IV as one of the lightest, faintest and farthest satellites ever seen ^[3].

Early Data Preview 2 covered roughly 7% of the night sky, collected between April 2025 and January 2026, about ten months of observing ^[8][4]. One new satellite in a fourteenth of the sky scales to something like 14 across the whole sky at the same depth ^[1]. The full survey will go deeper than that: it takes repeat images every few nights, and Rubin's car-sized camera can detect objects up to 100 million times dimmer than the naked eye can see ^[7]. Converting detections into a population also needs a detection efficiency, the fraction of real ultrafaint dwarfs at a given brightness and distance that the pipeline recovers. The research note does not report one.