Julian Shapiro, a first-year Berkeley student, has identified three isolated dwarf galaxies that may have been flung out of the Pinwheel Galaxy, using public telescope archives he searched while still at high school.
Three Quiet Dwarf Galaxies Near the Pinwheel
The paper, published in The Astrophysical Journal on 6 October 2026, reports no sign of recent star formation in ultraviolet images of Shapiro Dwarf Galaxies I, II and III. They lie roughly 940,000 to 1.07 million light-years from M101, the Pinwheel Galaxy, in projection; that spiral is about 22 million light-years from Earth.
Backsplash galaxies are dwarfs that swing through a larger galaxy's halo, lose their gas to its pressure and are thrown back out. Shapiro, the paper's sole author, proposes that M101 did this to all three, while the paper keeps a rival explanation open for one of them.
The standard cosmological model, lambda-cold dark matter or ΛCDM [INT-4], predicts that such galaxies should be common, yet clear local examples have been elusive, according to UC Berkeley.
The paper points to larger confirmed samples as the route to consistency tests of the model, while Berkeley's release says the findings suggest it holds in milder cosmic neighbourhoods.
Subaru Archives and a Tenerife Telescope Built the Case
Imagine hunting for galaxies that astronomers have predicted but never confirmed. You would ask simulations where they should hide, then search images other people had already taken. Berkeley says that is broadly what Shapiro did.
DG-I and DG-II turned up in archival Hyper Suprime-Cam data from the Subaru Telescope on Maunakea, and DG-III in that camera's legacy archive, the paper says. Shapiro says he developed his own modelling and analysis code, and learned the techniques partly from journals that are free to read.
The one fresh dataset was four nights of hydrogen-alpha imaging of DG-I with an 80-centimetre (31-inch) telescope in the Two-meter Twin Telescope facility on Tenerife. Those exposures found no hydrogen-alpha emission from the galaxy, which would signal recent star formation.
Radio data from the Westerbork Synthesis Radio Telescope in the Netherlands also showed no hydrogen gas at DG-I or DG-II, down to limits of roughly 140,000 and 300,000 solar masses.
Hubble and Webb Follow-Up Is Needed for All Three
Ask what would change your mind. The paper calls all three galaxies candidates and treats their distances as provisional: about 13.7 million light-years for DG-I, 20.2 million for DG-II and 26.1 million for DG-III, each with a wide margin of error.
DG-II lies inside the projected halo of the spiral NGC 5585 and may be its faint satellite instead, and stripping by dense filaments of the cosmic web is not ruled out. DG-I was found at about the same time by the ID-MAGE survey, which the paper credits.
Hubble or Webb [INT-3] imaging would fix distances from red giant stars, and Keck spectra would measure velocities, the paper says. It also notes a 2025 Hubble study of six isolated, quiescent Local Group dwarfs that found none likely to be backsplash galaxies within 6 billion years, with Cetus the one plausible case.
Shapiro expects the Vera C. Rubin Observatory [INT-1], whose 10-year survey of the southern sky began on 29 June 2026, to reveal many more candidates, even and especially among dwarfs this faint. UK institutions take part through the LSST:UK consortium.
Shapiro is now in his first semester at Berkeley, where he will join astronomy professor Alex Filippenko's group on the Hubble tension [INT-2] and hopes to run follow-up observations on the galaxies. He says he wants the work to inspire other young people, arguing that public archives let them contribute to real discoveries.
Filippenko called his school research 'remarkably sophisticated' in the Berkeley release, which also notes Shapiro is a 2026 Davidson Fellow and won the Astronomical League's National Young Astronomer Award in 2025.
