About CV Research Posters Outreach

I am an NSF Astronomy & Astrophysics Postdoctoral Fellow at the Harvard-Smithsonian Center for Astrophysics. I am also a research associate at the American Museum of Natural History in New York.

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.

(download high-quality PDF)

In my PhD thesis, completed in 2013 under the supervision of Profs. Dan Maoz (TAU) and Michael Shara (AMNH), I measured the rates at which Type Ia supernovae exploeded in three surveys, using both imaging and spectroscopy, and showed that these rates were consistent with a progenitor composed of two carbon-oxygen white dwarfs. In 2012, my thesis was animated by Jorge Cham for PhD Comics:

Besides research, I am a AAS Astronomy Ambassador and the director of the Science Research Mentoring Program at Harvard.

In my free time, I write fiction, and read as much as I can. I am also interested in the history, politics, and culture of ancient Rome, political science, ornithology, and experiencing the world by living somewhere else every once in a while.

Curriculum Vitae
Publications [ ADS | astro-ph ]
Ph.D. thesis  [ ADS | pdf ]

Perkin 317
Harvard-SAO Center for Astrophysics
60 Garden Street, MS-20
Cambridge, MA 02138

Voice: (617) 495-7733
E-Mail: or.graur @ cfa.harvard.edu

Science Research Mentoring Program


Current highlight
In Graur et al. (2019), 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, observed with the Hubble Space Telescope
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 Graur et al. (2016), we used SN 2012cg to show for the first time that the late-time light curve significantly deviated from pure 56Co decay. Watch video (left) for details.

In Graur et al. (2018b), we presented late-time observations of SN 2015F. Together with SNe 2011fe, 2012cg, and 2014J, our measurements suggest a correlation between the intrinsic luminosity of SNe Ia and the degree to which the light curve deviates from pure 56Co >900 days after explosion. This correlation was then strengthened by the addition of ASASSN-14lp in Graur et al. (2018c) and new observations of SN 2014J in Graur (2019).

In Graur et al. (2019), we show that the near-infrared light curves of Type Ia supernovae plateau between 150 and 500 days after explosion.

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.

Pre-explosion imaging of Type Ia supernovae with the Hubble Space Telescope

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).


Click on any of the posters for the full-resolution version.

Course Summaries and Data Analysis Handbook

Feel free to download the following summaries from my B.Sc. in Physics, but be advised that they are all in Hebrew. The Data Analysis Handbook was written for 1st-year Physics Lab students at Tel-Aviv University, and is also used at the Hebrew University in Jerusalem.

Data Analysis for 1st-year Undergraduate Physics Lab
Also available from the Tel-Aviv University Neiman Library of Exact Sciences

Course summaries: 1st year | 2nd year | 3rd year

Education and Outreach

As I describe in Graur (2018a), education and public outreach are an integral part of my career as a scientist. I have been involved in many outreach programs, including the Science Research Mentoring Program (Graur 2018b).

The Science Research Mentoring Program

I am the director of the Science Research Mentoring Program (SRMP) at the Harvard-Smithsonian Center for Astrophysics. High-school students are paired with Harvard graduate students and postdocs who advise them in year-long, independent research projects. The program is described in detail in Graur (2018b). The first cohort of ten high-school students from the Cambridge Rindge and Latin School graduated from the program in May 2018; the 2018-2019 cohort numbers 11 students. The program also runs a hands-on lecture series at the high school's Aerosapce Engineering and Astronomy Club. The syllabi of the activities are available from this website.

During 2011-2016, I was a mentor in the Science Research Mentoring Program at the American Museum of Natural History. Each year, I mentored 3-4 high-school students by giving them a research project that was part of my wider work. My students have searched for supernovae in Hubble Space Telescope images, digitized catalogs of variable stars in the Magellanic Clouds, and studied the evolving periodicity of Cepheids.

Some of my students have been featured in Shelf Life, a web-series about the collections of the museum. Two of my students presented their work in the 225th meeting of the American Astronomical Society in Seattle, WA in 2015.

American Museum of Natural History Science Bulletins During 2014-2016, I was the science advisor for the astronomy Science Bulletins produced by the American Museum of Natural History, in which we report on current research in astrophysics.

During my graduate studies at Tel Aviv University, I was one of the organizers of the Tel Aviv University AstroClub, an outreach unit led by graduate students that organized physics and astronomy talks and events for the general public.

Webpage designed with the kind help of Keren Sharon. Images courtesy of NASA, ESA, SDSS, Subaru, and the Lick Observatory. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation.