Training Custom Light Curve Models of SN Ia Sub-Populations Selected According to Host Galaxy Properties

G. Taylor,C. Lidman,B. Popovic,H. Abbot

2024-01-14

Abstract:Type Ia supernova (SN Ia) cosmology analyses include a luminosity step function in their distance standardization process to account for an observed yet unexplained difference in the post-standardization luminosities of SNe Ia originating from different host galaxy populations (e.g., high-mass ($M \gtrsim 10^{10} M_{\odot}$) versus low-mass galaxies). We present a novel method for including host-mass correlations in the SALT3 light curve model used for standardising SN Ia distances. We split the SALT3 training sample according to host-mass, training independent models for the low- and high-host-mass samples. Our models indicate that there are different average Si II spectral feature strengths between the two populations, and that the average SED of SNe from low-mass galaxies is bluer than the high-mass counterpart. We then use our trained models to perform a SN cosmology analysis on the 3-year spectroscopically confirmed Dark Energy Survey SN sample, treating SNe from low- and high-mass host galaxies as separate populations throughout. We find that our mass-split models reduce the Hubble residual scatter in the sample, albeit at a low statistical significance. We do find a reduction in the mass-correlated luminosity step but conclude that this arises from the model-dependent re-definition of the fiducial SN absolute magnitude rather than the models themselves. Our results stress the importance of adopting a standard definition of the SN parameters ($x_0, x_1, c$) in order to extract the most value out of the light curve modelling tools that are currently available and to correctly interpret results that are fit with different models.

Cosmology and Nongalactic Astrophysics,Astrophysics of Galaxies,High Energy Astrophysical Phenomena

What problem does this paper attempt to address?

### The problems the paper attempts to solve

This paper attempts to solve the "mass step" problem in the standard candle analysis of Type Ia supernovae (SN Ia). Specifically, the researchers have observed that Type Ia supernovae from host galaxies of different masses have differences in their standardized brightness. This difference is usually dealt with by introducing a brightness correction term, but its physical origin has not been fully understood. The paper proposes a new method to more accurately explain this phenomenon by incorporating the correlation of host galaxy masses into the SALT3 light - curve model.

### Background and motivation

Type Ia supernovae are widely used to measure cosmic distances because of their stable brightness, especially in the study of the history of cosmic expansion and dark energy. However, recent studies have found that Type Ia supernovae from host galaxies of different masses have significant differences in their standardized brightness, which is known as the "mass step". This difference may be related to factors such as the metallicity and star - formation rate of the host galaxy, but the exact cause is still unclear.

### Research methods

1. **Data set**:
- Used 1083 Type Ia supernovae in the K21 sample and divided them into two sub - samples, "high - host - mass" (> 10 ^ 10 solar masses) and "low - host - mass" (< 10 ^ 10 solar masses) according to the mass of the host galaxy.
- Supplemented data from SNLS and JLA to obtain complete host galaxy mass information.

2. **Model training**:
- Used the SALT3 model to train these two sub - samples separately to generate independent SALT3.HIGHMASS and SALT3.LOWMASS models.
- Adjusted the training parameters to ensure the coverage of the model in different wavebands and phases.

3. **Result evaluation**:
- Used the 3 - year spectroscopically confirmed Type Ia supernovae sample of the Dark Energy Survey (DES - SN3YR) to fit the light - curves with the SALT3.HIGHMASS and SALT3.LOWMASS models respectively.
- Evaluated the performance of these models in cosmological analysis, especially how they affect the dispersion of Hubble residuals.

### Main findings

- **Model performance**:
- The SALT3.HIGHMASS and SALT3.LOWMASS models show better fitting effects on supernovae with high - host - mass and low - host - mass respectively.
- These models reduce the dispersion of Hubble residuals in the sample, although the statistical significance is low.

- **Physical explanation**:
- The models indicate that Type Ia supernovae with high - host - mass and low - host - mass have significant differences in the average Si II spectral feature intensity and the average SED.
- The spectral energy level distribution of Type Ia supernovae with low - host - mass is bluer than that with high - host - mass.

- **Cosmological impact**:
- When using these models for cosmological analysis, the influence of the mass step is reduced, but this is mainly because the models re - define the absolute magnitude of the reference SN, rather than the change of the models themselves.

### Conclusion

This study emphasizes the importance of adopting a unified SN parameter definition in light - curve modeling tools in order to extract the maximum value from existing models and correctly interpret the results fitted with different models. Although the mass step problem has not been completely solved, this method provides a new perspective and helps to better understand the properties of Type Ia supernovae and their applications in cosmology.