Type Ia Supernovae Could Use Some More Color

Title: Dependence of Type Ia supernova luminosities on their local environment

Authors: M. Roman, D. Hardin, M. Betoule, P. Astier, C. Balland, R. S. Ellis, S. Fabbro, J. Guy, I. Hook, D. A. Howell, C. Lidman, A. Mitra, A. Möller, A. M. Mourão, J. Neveu, N. Palanque-Delabrouille, C. J. Pritchet, N. Regnault, V. Ruhlmann-Kleider, C. Saunders, M. Sullivan

First Author’s Institution: Sorbonne Université, Université Paris Diderot, CNRS/IN2P3, Laboratoire de Physique Nucléaire et de Hautes Energies, LPNHE, Paris, France

Status: Published in Astronomy & Astrophysics; open access on Arxiv

Type Ia supernovae (SNe Ia) are brilliant stellar explosions that are very important for cosmology. Their standardized luminosities are one of the primary tools that we can use to calculate how fast the universe is expanding! However, SNe Ia do not come standardized. There is variation among their observed properties, and these differences must be corrected through empirical relationships in order to use SNe Ia as distance indicators. In today’s paper, the authors explore whether the properties of the supernova host galaxies can yield another empirical relationship to better standardize SNe Ia. In order to understand how this can help us, let’s start with a quick introduction into how we use SNe Ia right now.

Type Ia Supernovae

There are two primary channels for the production of SNe Ia: either two white dwarfs merge and explode, or a single white dwarf accretes matter from a companion star until it reaches the Chandrasekhar Limit. At this point, the white dwarf undergoes a runaway nuclear reaction and explodes. These channels and their predicted observational signatures can give us clues about stellar evolution, and they are discussed in detail in this astrobite from 2012 and this one from 2015.

Figure 1: Light curves for many different SNe Ia are plotted on the left. A brightness correction is made for each one based on how fast its light curve declines in time on the left, leading to a relatively standard graph of luminosity vs time for all SNe Ia on the right. (Figure taken from Durham University Department of Physics)

Regardless of their production mechanism, SNe Ia also happen to be standardizable candles. By comparing the brightness we observe to their standardized luminosities, we can figure out how far away SNe Ia are from us. We then can look at how much their light is redshifted and determine how quickly these objects are moving away from us. Coupling this distance and velocity information allows us to calculate exactly how fast the universe is expanding (spoiler: it’s accelerating!). This very calculation led to the idea of dark energy and the Nobel Prize in Physics in 2011.

However, not all SNe Ia are created equal. There is enough variation among the explosions to complicate their use as distance indicators. Figure 1 shows some of this variation with regards to peak brightness and how quickly different SNe Ia light curves decline. This is the reason they are standardizable candles instead of standard candles. Thankfully, we can use what we know about how SNe Ia behave to correct for this variation we observe, allowing us to standardize their luminosities – billions of times as luminous as the sun!  

Corrections with a Splash of Color

Today’s paper suggests that a new correction needs to be added to our distance calculations using SNe Ia light curves. Normal corrections applied to SNe Ia are based on two factors: stretch, which is how fast the explosion declines in brightness; and color, which is essentially B-V color at peak brightness. These corrections enable us to get pretty consistent standardized luminosities for SNe Ia; the largest modern samples have scatters of only around 1% or so at peak brightness. However, the authors of today’s paper suggest that another correction for the color of the local environments of the supernovae should be included. In other words, they believe that we should not just consider the supernova itself – we should also consider its surroundings. Examples of local regions that are under consideration are shown in Figure 2.

Figure 2: Examples of galaxies and the local environments of supernovae. In each image, the galaxy is enclosed by the dashed line and the local environment of the supernova is encircled by the solid line. The supernova occurred at the red cross inside the solid circle. The authors of the paper suggest that we should also consider the region inside the solid circle when applying corrections, rather than just the supernovae itself. (Figure cropped and taken from Fig. 2 in the paper)

The motivation for this hypothesis comes from the relationship between a galaxy’s color and its star formation rate. Redder galaxies are not forming many stars, while bluer galaxies have much more active star formation. Previous analyses have found that galaxies that are not forming many stars tend to host SNe Ia that are brighter than SNe Ia from very active galaxies even after standardization.

The authors test their hypothesis by looking for a correlation between Hubble residuals and local color. Hubble residuals are the differences between our best fit cosmological model and the SNe Ia data on a distance vs. cosmological redshift plot (known as a Hubble diagram). If there is a correlation between these residuals and the color of the supernova’s surroundings, it means that adding color to our model could help improve the distance estimates to SNe Ia.

Evidence of the Need for a New Correction

Figure 3: Plot of Hubble residuals (y-axis) vs. local U-V color (x-axis) in magnitudes. Points are essentially colored by redshift – the green points are the closest, the red are at medium redshift, and the blue are at the farthest. The blue histogram on the right corresponds to the points left of the dashed line, and the red histogram corresponds to the points right of it. The purple dashed line shows the apparent correlation between Hubble residuals and local U-V color, indicating we should correct for this parameter in our model. (Figure taken Fig. 13 in the paper)

Figure 3 compares Hubble residuals to the local U-V color. The authors perform a statistical analysis and find that there is only a 1 in 400 billion chance (7 sigma) of having no correlation between the Hubble residuals and local U-V color. This strongly suggests that local U-V color should be used to correct SNe Ia light curves. This would presumably improve our model and reduce the scatter in our sample.

The major takeaway from the paper is that we should consider both the properties of the supernova explosion and its surroundings in our cosmological analyses. With major observational projects (LSST and WFIRST) planned for the coming decades that will help us observe more SNe Ia than ever before, corrections like these will help us determine the expansion of the universe more accurately and, consequently, could help us finally pin down the nature of dark energy!


About Michael Foley

I'm a graduate student studying Astrophysics at Harvard University. My research focuses on using simulations and observations to study stellar feedback - the effects of the light and matter ejected by stars into their surroundings. I'm interested in learning how these effects can influence further star and galaxy formation and evolution. Outside of research, I'm really passionate about education, music, and free food.


  1. Congrats on your first post, Mike! This post actually came at the perfect moment for me as I am writing a talk on the cosmic distance ladder and needed information on the Type Ia SNe, so thanks! 😀 Hope you are doing well and your studies are going well! Abi Chown 🙂

    • Thanks, Abi! Really glad to hear you found it helpful. Let me know if you need any more information. Hope all is going well for you too! Good luck on your talk!



  1. Daily Study Log (2019-02-18) | Study Astrophysics - […] have read the Astrobites article. This article indicates that the SNe Ia should be corrected by using the local U-V…

Leave a Reply