Research

I study different types of transients – astrophysical phenomena that change on human timescales. I mostly work on supernovae – the explosions of stars – and try to figure out which stars explode as different types of supernovae. To do so, I use images and spectra of the supernovae and their host galaxies, taken with ground and space-based telescopes. My “Supernova Tree of Death,” below, shows this in graphical form.

(click on the image to download a high-quality PDF file)

Current highlight

In Graur et al. (2020), I show that the near-infrared light curves of Type Ia supernovae plateau between 150 and 500 days after explosion (see figure below). Only Type IIP (where the P stands for “plateau”) have been known to display light curve plateaus. There, the plateau is due to a coincindence between the expansion of the supernova ejecta and the recombination timescale of hydrogen. Type Ia supernovae, by definition, are devoid of hydrogen, so there must be a different reason for the plateau. One option is scattering of UV photons from iron lines in the UV into the optical and near-infrared. This scattering might even explain why, instead of crashing, the optical light curves continue to decline and even slow down (Graur et al. 2016, 2018b, 2018c; Graur 2019). Read the full Nature Astronomy article for more.

Supernova rates and the delay-time distribution
The rates at which different types of supernovae explode in various environments can help us constrain the models proposed for their progenitors. Rates as a function of redshift show us how the supernova rate changed over cosmic time. When examining a specific sample of galaxies, correlations between the rates and these galaxies’ properties (their stellar mass, star-formation rates, or metallicities) can also connect us to the progenitors of these explosions. I use the rates of Type Ia supernovae to reconstruct their delay-time distribution (DTD), a strong diagnostic of progenitor scenarios. The most up-to-date collection of rates and DTD measurements can be found in Maoz & Graur (2017).

We re-analyzed the landmark LOSS supernova sample and measured new absolute and relative rates. The rates of stripped-envelope supernovae are consistent with binary progenitors, but now with single-star progenitors. The relative rates of Type Ia supernovae in low-mass and high-mass galaxies are consistent with binary white-dwarf progenitors. Select papers: Shivvers et al. (2017), Graur et al. (2017a), Graur et al. (2017b), Shen, Toonen, & Graur (2017), Chakrabarti et al. (2018).

Supernovae – and other transients – can be discovered in large-scale spectroscopic galaxy surveys. In some of these galaxies, supernovae will serendipitously explode in the area covered by the spectral aperture. It is only a matter of prying the supernova signal from that of the underlying galaxy. Select papers: Graur & Maoz (2013), Graur et al. (2015).

Using the Hubble Space Telescope, we searched for supernovae in and behind galaxy clusters, as well as in “regular” galaxy fields. Select papers: Graur et al. (2014a) (CLASH Ia rates), Rodney et al. (2014) (CANDELS Ia rates), Strolger et al. (2015) (core-collapse rates), Riess et al. (2018) (cosmology).

Using the 8.2-m Subaru Telescope, we measured Type Ia supernova rates out to redshift 2. To date, this is the only ground-based survey to reach this high redshift. Select papers: Graur et al. (2011) (press release), Finkelman, Graur, & Brosch (2011).

The late-time light curves of Type Ia supernovae
The way Type Ia supernovae continue to fade >900 days after explosion (when the supernovae are a million times fainter than at peak!) can be used as a new diagnostic of nebular physics, as well as progenitor and explosion scenarios.

In a series of papers (Graur et al. 2016, 2018b, 2018c, 2019, 2020), I show that the light curves of Type Ia supernovae significantly deviate from pure 56Co decay, that there may be a correlation between how much the light curve slows down and the intrinsic luminosity of the supernovae, and that there is a year-long plateau phase in the near-infrared. Watch video (below) for details.

Tidal Disruption Events and their host galaxies
When a star strays too close to the super-massive black hole at the center of its galaxy, it can be disrupted by the black hole’s gravitational pull. Half of the star’s material will be flung away and half will fall into the black hole, giving rise to a luminous flare. The rates of these tidal disruption events (TDEs) can help us understand how they come about, as well as probe the conditions in the direct vicinity of the black hole.

Using public spectra of 35 TDE host galaxies, in Graur et al. (2018a) we showed that TDEs explode in all types of galaxies, but they prefer dense galaxies. We devise a paramteric model that relates the TDE rate to global galaxy properties, and fit it to our data. The results are consistent with the scenario where the TDE rate is set by the dynamical relaxation of the stars in the vicinity of the black hole.

Type Ia supernova archaeology
With a little luck, images of the area where a supernova explodes will already exist before the explosion, allowing us to identify and study the star before it exploded. In Graur et al. (2014b), we used pre-explosion HeII images of SN 2011fe (which exploded in M101, the “pinwheel” galaxy) to constrain the “single-degenerate” progenitor scenario. A similar experiment on SN 2014J found no evidence of the expected emission-line nebula in either [O III] or Hβ (Graur & Woods 2019).

Images courtesy of NASA, ESA, SDSS, Subaru, and the Lick Observatory.