Finding the Fuse: Prospects for the Detection and Characterization of Hydrogen-Rich Core-Collapse Supernova Precursor Emission with the LSST

The Big Picture

For decades, astronomers watched stars explode and wondered whether there were warning signs they’d missed. A few were caught brightening before the final blast. With a powerful new sky survey about to come online, we’re going to find out how common that really is.

Massive stars, those at least eight times heavier than our Sun, end their lives in explosions called supernovae. But a growing number of these stellar deaths don’t come without warning. Several supernovae have shown bursts of energy in the months or years before detonation, pointing to violent instabilities in a star’s final stages. These precursor eruptions, flares of light that precede the main explosion by days to years, let us peer into what’s happening inside a star as it tears itself apart.

The catch: we’ve only spotted a handful. A new paper from IAIFI researchers at Harvard, MIT, and collaborators worldwide uses detailed simulations to forecast how many precursors the upcoming Vera C. Rubin Observatory will detect and what it will take to find them.

Key Insight: The Rubin Observatory’s LSST survey could detect between 150 and 400 supernova precursor events in just its first three years, turning a rare curiosity into a statistical sample large enough to test theories of stellar death.

How It Works

The Legacy Survey of Space and Time (LSST) will photograph the entire southern sky every few nights for a decade, building an enormous archive of how objects change over time. That cadence is exactly what’s needed to hunt for supernova precursors, which can flicker weeks or months before the main explosion.

The researchers built their forecasts by simulating realistic precursor light curves (records of how a star’s brightness changes over time) drawn from observed events and theoretical models. They anchored the work on real detections:

• SN 2020tlf, a hydrogen-rich supernova whose dying star showed 130 days of plateau-like brightening before exploding
• SN 2010mc, a Type IIn supernova with a dramatic eruption roughly 40 days before the main event
• SN 2015bh, which showed multi-year variability before its terminal explosion
• Several theoretical models for hydrogen recombination-driven outbursts

A pattern runs through many of these detections: the precursors show long-lived, plateau-like light curves rather than sharp spikes. This is the signature of hydrogen recombination, where superheated hydrogen gas gradually cools and releases energy steadily over weeks. Think of a lava flow glowing for days after a volcanic eruption, not just at the moment of the blast.

The team tested two complementary detection strategies. The first searched single-epoch observations: individual nightly images where a bright enough precursor triggers an alert. The second used binned photometry, stacking many faint measurements over 20- to 100-day windows to reveal light too dim for any single night’s image.

For single-epoch searches, detection rates run roughly 40 to 130 precursors per year for Type IIP/IIL supernovae (the most common hydrogen-rich variety, named for the characteristic shapes of their brightness curves) and around 110 per year for the more extreme Type IIn class, named for the narrow spectral fingerprints of slow-moving gas surrounding the star before explosion. Those numbers already dwarf the handful of precursors known today.

Stacking images is where the real payoff comes in. Over LSST’s first three years, binning in optimally sized windows yields between 150 and 400 total detections. The right bin size tracks the underlying physics: 100-day bins work best for long-lived plateau-like precursors similar to SN 2020tlf, while 20-day bins are better tuned to shorter, sharper eruptions. Choosing the bin size is a bit like choosing the right shutter speed for a photograph.

The paper also tackles template contamination. To detect a precursor, astronomers subtract a reference image of the host galaxy from new observations. If a precursor began months before the reference was taken, its light is already baked into the baseline, hiding subsequent brightening. The team quantifies how much this degrades detection rates and proposes strategies to estimate and subtract the residual flux. Getting this right will be essential for making LSST searches work in practice.

Why It Matters

Precursor eruptions are one of the most direct probes of what massive stars do in their final years. We still lack a clear picture of this period. Why do some stars erupt so violently before death?

Leading candidates include convective wave damping, where energy from deep nuclear burning propagates outward and shakes the star’s outer layers. There are also pulsation-driven superwinds, rhythmic pulsations that fling material from the star’s surface into space. And there’s binary star interaction, where a companion’s gravity triggers instability. Each mechanism would leave different imprints on a precursor’s duration, brightness, and color.

With hundreds of detections rather than a handful, LSST will give researchers enough events to test these predictions against real data. Splitting a precursor’s light to identify its chemical fingerprint and velocity will let astronomers measure the composition of ejected material and pin down mass-loss rates.

Rubin Observatory’s survey is set to begin operations soon. This study is a preparation paper: a plan for designing searches, setting detection thresholds, and handling data challenges before the data starts pouring in. Getting the strategy right now means not missing hundreds of cosmic death-rattles once LSST is live.

Bottom Line: By simulating the full range of known precursor models against realistic LSST observing conditions, this study shows the coming survey will vastly expand our ability to catch stars in the act of dying, potentially detecting hundreds of pre-explosion flares and pinning down the physics of how massive stars shed mass in their final years.