Change page style: 

Game Over for Supernovae Hide & Seek

January 12, 2018

SN 2013if with GeMS/GSAOI, from left to right with linear scaling: Reference image (June 2015), discovery image (April 2013) and the image subtraction. SN 2013if had a projected distance from the nucleus as small as 600 light years (200 pc), which makes it the second most nuclear CCSN discovery in a LIRG to date in the optical and near-IR after SN 2010cu.

The Core-collapse Supernova Rate Problem, or the fact that we don’t see as many core-collapse supernovae as we would expect, has a solution, thanks to research using the Gemini South telescope. The research team concludes that the majority of core collapse supernovae, exploding in luminous infrared galaxies, have previously not been found due to dust obscuration and poor spatial resolution.

Core-collapse supernovae are spectacular explosions that mark the violent deaths of massive stars. An international team of astronomers, led by PhD student Erik Kool of Macquarie University in Australia, used laser guide star imaging on the Gemini South telescope to study why we don’t see as many of these core-collapse supernovae as expected. The study began in 2015 with the Supernova UNmasked By InfraRed detection (SUNBIRD) project which has shown that dust obscuration and limited spatial resolution can explain the small number of detections to date.

In this, the first results of the SUNBIRD project, the team discovered three core-collapse supernovae, and one possible supernova that could not be confirmed with subsequent imaging. Remarkably, these supernovae were spotted as close as 600 light years from the bright nuclear regions of these galaxies – despite being at least 150 million light years from the Earth. “Because we observed in the near-infrared, the supernovae are less affected by dust extinction compared to optical light,” said Kool.

According to Kool the results coming from SUNBIRD reveal that their new approach provides a powerful tool for uncovering core-collapse supernova in nuclear regions of galaxies. They also conclude that this methodology is crucial in characterizing these supernova that are invisible through other means. Kool adds, “The supernova rate problem can be resolved using the unique multi-conjugate adaptive optics capability provided by Gemini, which allows us to achieve the highest spatial resolution in order to probe very close to the nuclear regions of galaxies.” This work is published in the Monthly Notices of the Royal Astronomical Society.

This research is also highlighted in the January 2018 GeminiFocus (p.11).


Core collapse supernova (CCSN) rates suffer from large uncertainties as many CCSNe exploding in regions of bright background emission and significant dust extinction remain unobserved. Such a shortfall is particularly prominent in luminous infrared galaxies (LIRGs), which have high star formation (and thus CCSN) rates and host bright and crowded nuclear regions, where large extinctions and reduced search detection efficiency likely lead to a significant fraction of CCSNe remaining undiscovered. We present the first results of project SUNBIRD (Supernovae UNmasked By InfraRed Detection), where we aim to uncover CCSNe that otherwise would remain hidden in the complex nuclear regions of LIRGs, and in this way improve the constraints on the fraction that is missed by optical seeing-limited surveys. We observe in the near-infrared 2.15 µm Ks-band, which is less affected by dust extinction compared to the optical, using the multi-conjugate adaptive optics imager GeMS/GSAOI on Gemini South, allowing us to achieve a spatial resolution that lets us probe close in to the nuclear regions. During our pilot program and subsequent first full year we have discovered three CCSNe and one candidate with projected nuclear offsets as small as 200 pc. When compared to the total sample of LIRG CCSNe discovered in the near-IR and optical, we show that our method is singularly effective in uncovering CCSNe in nuclear regions and we conclude that the majority of CCSNe exploding in LIRGs are not detected as a result of dust obscuration and poor spatial resolution.