A Probabilistic Autoencoder for Type Ia Supernova Spectral Time Series

Journal Article · Mon Aug 08 00:00:00 EDT 2022 · The Astrophysical Journal

DOI:https://doi.org/10.3847/1538-4357/ac7c08· OSTI ID:1907591

Stein, George ^[1]; Seljak, Uroš ^[1]; Böhm, Vanessa ^[1]; Aldering, G. ^[2]; Antilogus, P. ^[3]; Aragon, C. ^[4]; Bailey, S. ^[2]; Baltay, C. ^[5]; Bongard, S. ^[3]; Boone, K. ^[6]; Buton, C. ^[7]; Copin, Y. ^[7]; Dixon, S. ^[8]; Fouchez, D. ^[9]; Gangler, E. ^[10]; Gupta, R. ^[2]; Hayden, B. ^[11]; Hillebrandt, W. ^[12]; Karmen, M. ^[2]; Kim, A. G. ^[2] ; Kowalski, M. ^[13]; Küsters, D. ^[14]; Léget, P. -F. ^[3]; Mondon, F. ^[15]; Nordin, J. ^[16]; Pain, R. ^[3]; Pecontal, E. ^[17]; Pereira, R. ^[7]; Perlmutter, S. ^[8]; Ponder, K. A. ^[18]; Rabinowitz, D. ^[5]; Rigault, M. ^[7]; Rubin, D. ^[19]; Runge, K. ^[2]; Saunders, C. ^[20]; Smadja, G. ^[7]; Suzuki, N. ^[21]; Tao, C. ^[22]; Taubenberger, S. ^[12]; Thomas, R. C. ^[2]; Vincenzi, M. ^[2]

---

1. University of California, Berkeley, CA (United States); Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States)
2. Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States)
3. Sorbonne University, Paris (France)
4. Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States); University of Washington, Seattle, WA (United States)
5. Yale University, New Haven, CT (United States)
6. Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States); University of California, Berkeley, CA (United States); University of Washington, Seattle, WA (United States)
7. Université Claude-Bernard Lyon 1 (UCBL) (France)
8. Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States); University of California, Berkeley, CA (United States)
9. Aix-Marseille University, Marseille (France)
10. Université Claude-Bernard Lyon 1 (UCBL) (France); University Clermont Auvergne, Clermont-Ferrand (France)
11. Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States); Space Telescope Science Institute, Baltimore, MD (United States)
12. Max-Planck-Institut für Astrophysik, Garching (Germany)
13. Humboldt University of Berlin (Germany); German Electron Synchrotron DESY, Hamburg (Germany)
14. University of California, Berkeley, CA (United States); German Electron Synchrotron DESY, Hamburg (Germany)
15. University Clermont Auvergne, Clermont-Ferrand (France)
16. Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States); Humboldt University of Berlin (Germany)
17. University of Lyon (France)
18. University of California, Berkeley, CA (United States)
19. Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States); University of Hawaii, Honolulu, HI (United States)
20. Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States); University of California, Berkeley, CA (United States); Princeton University, NJ (United States); Sorbonne University, Paris (France)
21. Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States); University of Tokyo (Japan). Kavli Institute for the Physics and Mathematics of the Universe (WPI)
22. Aix-Marseille University, Marseille (France); Tsinghua University, Beijing (China)

We construct a physically parameterized probabilistic autoencoder (PAE) to learn the intrinsic diversity of Type Ia supernovae (SNe Ia) from a sparse set of spectral time series. The PAE is a two-stage generative model, composed of an autoencoder that is interpreted probabilistically after training using a normalizing flow. We demonstrate that the PAE learns a low-dimensional latent space that captures the nonlinear range of features that exists within the population and can accurately model the spectral evolution of SNe Ia across the full range of wavelength and observation times directly from the data. By introducing a correlation penalty term and multistage training setup alongside our physically parameterized network, we show that intrinsic and extrinsic modes of variability can be separated during training, removing the need for the additional models to perform magnitude standardization. We then use our PAE in a number of downstream tasks on SNe Ia for increasingly precise cosmological analyses, including the automatic detection of SN outliers, the generation of samples consistent with the data distribution, and solving the inverse problem in the presence of noisy and incomplete data to constrain cosmological distance measurements. We find that the optimal number of intrinsic model parameters appears to be three, in line with previous studies, and show that we can standardize our test sample of SNe Ia with an rms of 0.091 ± 0.010 mag, which corresponds to 0.074 ± 0.010 mag if peculiar velocity contributions are removed.

• Research Organization: Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States)
• Sponsoring Organization: European Research Council (ERC); German Research Foundation (DFG); Gordon & Betty Moore Foundation; National Natural Science Foundation of China (NSFC); National Research Agency; Tsinghua University; USDOE Office of Science (SC), High Energy Physics (HEP)
• Grant/Contract Number: AC02-05CH11231
• OSTI ID: 1907591
• Journal Information: The Astrophysical Journal, Journal Name: The Astrophysical Journal Journal Issue: 1 Vol. 935; ISSN 0004-637X
• Publisher: IOP Publishing
• Country of Publication: United States
• Language: English

Related Subjects

79 ASTRONOMY AND ASTROPHYSICS