Imagine you're running a small bakery. You bake bread every day, but when you count the loaves at closing, half are missing. You know nobody ate them in-store. The only explanation: they walked out the door — carried away by wind, by customers, by forces you underestimated. That's what's happening in dwarf galaxies, except the bread is oxygen and the door is supernova-driven galactic winds. The GLOW (Galaxies Losing Oxygen via Winds) project, published in The Astrophysical Journal, used NASA's Hubble Space Telescope to study 37 nearby dwarf galaxies — systems containing a few thousand to hundreds of millions of stars. The team traced oxygen atoms from their birthplace in massive stars through the interstellar medium, dust, and potentially out into intergalactic space. The core finding: dwarf galaxies retain only a small fraction of the oxygen their stars produce, and the deficit is larger than current galaxy-evolution simulations predict. Oxygen serves as a tracer element here, not the endpoint. It's the most abundant element after hydrogen and helium, its production pathways are well understood, and it's relatively easy to detect spectroscopically. Tracking oxygen lets you reconstruct a galaxy's entire chemical enrichment history — how metals get made, where they go, and how much leaks into the intergalactic medium. The choice is pragmatic, not arbitrary. The study leveraged archival Hubble optical imaging to resolve individual stellar populations and reconstruct star-formation histories across the 37 targets. Additional data came from NASA's retired Spitzer Space Telescope (infrared), the Very Large Array, and the Australian Telescope Compact Array (both radio). The multi-wavelength approach let the team account for oxygen locked in stars, in interstellar gas, and in dust — leaving the residual as the amount expelled. The galaxies were selected specifically because their mass range falls where simulations predict escalating oxygen loss, and because they're close enough for Hubble to resolve individual stars. The mechanism driving the loss is galactic winds powered by supernovae. Massive stars burn fast, die explosively, and the blast energy in a low-mass galaxy — with its shallow gravitational well — can push enriched gas clear out of the system. The GLOW results indicate these winds are more efficient at stripping oxygen than current models account for. This matters for the intergalactic medium: if dwarf galaxies are bleeding metals at higher rates, the chemical enrichment of the space between galaxies is also underestimated. The study builds on a 2014 benchmark that addressed the same question for Milky Way-mass galaxies, finding significant oxygen loss even in larger systems. GLOW extends this to the low-mass regime across a wide range of masses and environments. A key open question remains: does the missing oxygen park in the circumgalactic medium — the diffuse halo around each galaxy — or does it escape entirely into the intergalactic medium? The current data can't fully resolve that. Future work is already mapped. Principal investigator Kristen Wingfield McQuinn has an approved program on the Nancy Grace Roman Space Telescope to find even lower-mass, fainter dwarf galaxies across wide sky areas, with follow-up planned using Hubble and JWST for deeper stellar archaeology. Co-investigator Grace Telford will use Hubble to zoom in on individual massive stars in GLOW targets to characterize how oxygen-producing stars differ in these small systems. Both programs aim to close the loop between stellar oxygen production and galactic-scale loss.
