AAS Nova

Shedding Light on Candles That Burn a Bit Too Bright

By on Astrobites
Share:
X-ray image of Kepler's Supernova remnant
Kepler's Supernova, the remnant of which is shown here in X-ray observations from the Chandra X-ray Observatory, is the most recent known Type Ia supernova in the Milky Way. It was discovered in 1604. [NASA/CXC/Univ of Texas at Arlington/M. Millard et al.]

Editor’s Note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org.

Title: 1991T-Like Type Ia Supernovae as an Extension of the Normal Population
Authors: John T. O’Brien et al.
First Author’s Institution: Michigan State University
Status: Published in ApJ

Famously, Type Ia supernovae have been used to measure the local Hubble constant, or the rate at which our universe expands. These objects earned the nickname “standard candles” since their near-constant intrinsic luminosities allow us to measure distances in space. Slowly but surely, however, we’ve learned that some of our standard candles aren’t that “standard” after all…

Branch classification diagram for Type Ia supernovae

Figure 1: Example of a “Branch classification” diagram for Type Ia supernovae. This figure compares the width of two silicon lines in Type Ia supernovae. Four groups are shown: shallow silicons (SS), broad lines (BL), cools (CL), and core normals (CN). Event SN 1991T is a member of the shallow silicons group (green triangles), indicating that the widths of the minor and major silicon lines are smaller than normal Type Ia supernovae (core normals). Click to enlarge. [Burrow et al. 2020]

Historically, Type Ia supernovae were proposed to develop from the transfer of mass between two stars, where the star receiving the mass was a carbon–oxygen white dwarf — the core of a low-mass to intermediate-mass star that’s reached the end of its life. After the white dwarf accretes a certain amount of mass, it explodes as a Type Ia supernova. Spectroscopic studies of these supernovae over the decades have shown a wide range of absorption features, one major absorption line being silicon, a key element produced in the explosion. In fact, a subclassification scheme of Type Ia supernovae — often referred to as the Branch classification — emerged based on the relative strengths of particular absorption features commonly identified in the spectra of these events (see Figure 1). One of these subclassifications is “shallow silicon,” which signifies a lack of silicon produced in the explosion. This subclassification (compared to other subclassifications in Figure 1) shows how Type Ia supernovae are like snowflakes: they have very similar structures yet vary in detail.

The supernova SN 1991T was the first observed event of its kind. What was so special about it? This event was considered over-luminous, or more luminous than the typical “near-intrinsic luminosity” of the average Type Ia supernova. Later, as observations improved, more events like SN 1991T were detected, contributing to the growing class of aptly named “1991T-like” events. The spectra of these events have shallow silicon lines compared to the normal range of Type Ia supernovae. The peculiarity of these absorption lines hints at something unique about these events, and the answer lies in studying the ejecta, or the ejected material in which chemical elements are produced. This article is a step toward understanding what differentiates these events from the norm and what we can infer about their origins.

Outside of this work, recent hydrodynamic simulations of various progenitor models, or stellar origins, have successfully recreated some of the observable signatures of Type Ia supernovae, including synthetic, or computed, optical spectra of theoretical events. Except, as previously mentioned, the observable signatures of Type Ia supernovae can vary quite a bit amongst all these subtypes and classifications! Instead of hydrodynamic simulations, the authors of this article chose to reconstruct the supernova ejecta using Bayesian inference and active learning conducted on early-time (within a few days after explosion) optical spectra of already observed normal and 1991T-like events. This is the time when 1991T-like events show their features! After training the model on this data, the authors developed a model to link the optical spectra and the ejecta properties corresponding to normal and 1991T-like events.

fraction of intermediate-mass elements versus ionization ratio

Figure 2: A plot showing the fraction of intermediate-mass elements (IME) as a function of the ionization ratio of the authors’ simulations. Moving to the right on the bottom axis indicates higher ionization states, whereas moving up on the left axis indicates more intermediate-mass elements for a given total ejecta mass. The break between blue stars (normal Type Ia supernovae) and orange stars (1991T-like supernovae) is called the “turnover.” Because the turnover is fairly smooth, it suggests that the progenitor, or stellar origin, of 1991T-like events might be similar to normal events. Click to enlarge. [O’Brien et al. 2024]

The team’s emulator successfully recreated both normal and 1991T-like events, at least with 68% confidence (think one sigma!). Furthermore, the authors discovered that the variety in the parameters used in their model illuminates some differences between these 1991T-like events and normal Type Ia supernovae. Remember those silicon features? They recreated those pesky absorption lines, particularly the major iron and silicon features experts look for. Their model successfully recreated silicon absorption features that were suppressed, or not as deep. This indicates a low fraction of intermediate-mass elements, which range from lithium to iron, produced in the explosion compared to the total mass. They also matched the deep, major iron line seen in 1991T-like events. Fewer intermediate-mass elements in 1991T-like supernovae suggest that these elements exist at higher ionization states than in normal Type Ia supernovae (see Figure 2). This suggests that there isn’t just a single mechanism that produces a 1991T-like supernova; it’s likely a combination of different physical processes.

The question now becomes: what can we learn about 1991T-like origins from this? Can a single progenitor model lead to different pathways? Or do we need different progenitor models to explain these differences in spectroscopic features? The authors believe fewer intermediate-mass elements and higher ionization states hint at normal and 1991T-like events sharing similar progenitor systems. In other words, 1991T-like events might just be an extension, or extreme, of the normal population. Perhaps the candle just burned a bit too bright!

Aside from this work, in addition to these over-luminous 1991T-like events, there also exists another interesting class of Type Ia supernovae dubbed “super-luminous,” which are roughly one, maybe two, magnitudes brighter than normal Type Ia supernovae. (Only in astronomy could the words over-luminous and super-luminous mean different things, right?) Because of this, researchers advocate for Type Ia supernovae to be called “standardizable” candles instead because, as you now know, their intrinsic luminosities really aren’t that uniform after all.

Original astrobite edited by Ansh Gupta and Dee Dunne.

About the author, Mckenzie Ferrari:

I’m a grad student at the University of Chicago. Most of my research focuses on simulations of Type Ia supernovae and galaxy formation and evolution.