An Extensive $\textit{Hubble Space Telescope}$ Study of the Offset and Host Light Distributions of Type I Superluminous Supernovae

The Big Picture

Imagine the most powerful explosions in the universe, blasts so bright they outshine entire galaxies for weeks. Now imagine that roughly 40% of them go off in the darkest, emptiest corners of their home galaxies, far from any obvious concentration of stars. That’s the puzzle at the heart of a new Hubble Space Telescope study of Type I superluminous supernovae (SLSNe), explosions that burn 10 to 100 times brighter than ordinary supernovae and stay visible for months.

Where a star explodes usually tells you something about where it lived. Massive stars that die young should blow up close to the star-forming regions where they were born. That logic holds for most supernovae. SLSNe keep breaking the pattern.

Brian Hsu, Peter Blanchard, Edo Berger, and Sebastian Gomez analyzed archival Hubble images of 65 SLSN host galaxies, spanning redshifts from z ≈ 0.05 to 2. It is the largest census yet of where these events occur within their hosts, and the results raise hard questions about what kind of star could produce them.

Key Insight: About 40% of all superluminous supernovae explode in the dimmest, most star-poor regions of their home galaxies, in sharp contrast to every other known type of massive-star death. Their parent stars may be runaways, flung from disrupted binary systems at hundreds of kilometers per second.

How It Works

The analysis hinges on Hubble’s angular resolution. Ground-based telescopes blur the fine structure of distant galaxies. Hubble can resolve individual star-forming clumps even at cosmological distances. The team used images from Hubble’s Advanced Camera for Surveys (ACS) and Wide Field Camera 3 (WFC3), all in rest-frame ultraviolet, the wavelength band that most directly traces young, hot, massive stars.

For each of the 65 SLSNe, they carried out precise astrometric matching, aligning Hubble images of each host galaxy against earlier images that caught the supernova mid-explosion. This pinpointed each blast to a specific pixel in the host galaxy image, often to within a fraction of an arcsecond.

From those positions, they measured three quantities:

• Physical offset: the distance in kiloparsecs between the supernova and its host galaxy’s center
• Host-normalized offset: that distance divided by the host’s half-light radius (the radius enclosing half the galaxy’s total brightness), allowing fair comparisons across galaxies of different sizes
• Fractional flux: the fraction of the host galaxy’s UV light coming from regions dimmer than the supernova’s location, a direct measure of how star-rich or star-poor that neighborhood is

Fractional flux is the most telling of the three. If a supernova lands on the brightest star-forming knot in its galaxy, fractional flux approaches 1: nearly all the galaxy’s light comes from regions fainter than that spot. A supernova just beyond the galaxy’s visible edge gets a fractional flux of 0.

Long gamma-ray bursts (LGRBs), brief intense flashes also produced by dying massive stars, cluster in the brightest regions of their hosts, with fractional fluxes skewed heavily toward 1. SLSNe look nothing like that. Roughly 40% of the sample sits at fractional flux = 0, in regions so faint they register essentially no UV light.

The normalized offsets tell the same story. SLSNe roughly follow an exponential disk profile, the expected brightness falloff if events tracked the galaxy’s starlight, but with a clear excess at large offsets of 1.5 to 4 times the host half-light radius. That excess is larger than what’s seen for LGRBs, and larger still compared to ordinary Type Ib/c and Type II core-collapse supernovae.

Why It Matters

The contrast with LGRBs is the sharpest result here. Long gamma-ray bursts are also hydrogen-stripped and energetic. They prefer the brightest, most actively star-forming regions of their hosts. For decades, SLSNe were lumped with LGRBs as products of similar stellar evolution: massive, fast-rotating, metal-poor stars that live fast and die spectacularly.

This analysis pulls them apart. The SLSN offset and fractional flux distributions are statistically distinct from those of LGRBs. Something is moving SLSN progenitors away from their birthplaces before they explode.

The favored explanation is runaway stars. Massive stars often form in binary pairs. When one star explodes, it can unbind the system and send the surviving companion racing through the galaxy at ~100 km/s. Over a typical SLSN progenitor lifetime of tens of millions of years, that’s enough to cover several kiloparsecs. The second explosion lands far from any star-forming region, sometimes in the dim outskirts or beyond the visible edge of the galaxy.

The team checked whether SLSN locations varied with redshift or with explosion parameters from the magnetar model, in which the blast is powered by the spin-down of an ultra-magnetic neutron star. No significant correlations turned up. If the runaway scenario is right, it holds across cosmic time.

If SLSNe come from runaway stars in disrupted binaries, the full evolutionary channel is tightly constrained: binary formation, first-star death, gravitational unbinding, and then a solitary drift across the galaxy before the second star collapses.

The Vera C. Rubin Observatory’s Legacy Survey of Space and Time will discover thousands of SLSNe in the coming years. That sample should be large enough to test the runaway hypothesis directly, by searching for velocity signatures and mapping SLSN environments across a wider range of host galaxy types.

Bottom Line: The largest Hubble study of superluminous supernova locations shows that these extreme explosions are systematically displaced from star-forming regions, unlike any other type of massive-star death. The most plausible explanation: their progenitors are massive runaway stars launched from disrupted binary systems.