Photometric Classification of 2315 Pan-STARRS1 Supernovae with Superphot

Hosseinzadeh, G.; Dauphin, Frederick; Villar, V. Ashley; Berger, E.; Jones, D. O.; Challis, Peter; Chornock, R.; Drout, M. R.; Foley, R. J.; Kirshner, R.; Lunnan, R.; Margutti, R.; Milisavljevic, Dan; Pan, Y.-C.; Rest, A.; Scolnic, Dan; Magnier, E. A.; Metcalfe, N.; Wainscoat, R. J.; Waters, C.

doi:10.3847/1538-4357/abc42b

Cited by 25 publications

(26 citation statements)

References 110 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our results are a significant improvement from both the SuperRAENN (Villar et al 2020) and Superphot (Hosseinzadeh et al 2020) models that have previously been applied to this dataset. We achieve an overall accuracy of 89% for the five-way classification compared to 87% for SuperRAENN on the same dataset.…”

Section: Photometric Classification On the Ps1 Datasetmentioning

confidence: 53%

ParSNIP: Generative Models of Transient Light Curves with Physics-Enabled Deep Learning

Boone

2021

Preprint

View full text Add to dashboard Cite

We present a novel method to produce empirical generative models of all kinds of astronomical transients from datasets of unlabeled light curves. Our hybrid model, that we call ParSNIP, uses a neural network to model the unknown intrinsic diversity of different transients and an explicit physics-based model of how light from the transient propagates through the universe and is observed. The ParSNIP model predicts the time-varying spectra of transients despite only being trained on photometric observations. With a three-dimensional intrinsic model, we are able to fit out-of-sample multiband light curves of many different kinds of transients with model uncertainties of 0.04-0.06 mag. The representation learned by the ParSNIP model is invariant to redshift, so it can be used to perform photometric classification of transients even with heavily biased training sets. Our classification techniques significantly outperform state-of-the-art methods on both simulated (PLAsTiCC) and real (PS1) datasets with 2.3× and 2× less contamination respectively for classification of Type Ia supernovae. We demonstrate how our model can identify previously-unobserved kinds of transients and produce a sample that is 90% pure. The ParSNIP model can also estimate distances to Type Ia supernovae in the PS1 dataset with an RMS of 0.150 ± 0.007 mag compared to 0.155 ± 0.008 mag for the SALT2 model on the same sample. We discuss how our model could be used to produce distance estimates for supernova cosmology without the need for explicit classification.

show abstract

Section: Photometric Classification On the Ps1 Datasetmentioning

confidence: 53%

ParSNIP: Generative Models of Transient Light Curves with Physics-Enabled Deep Learning

Boone

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…The data used in this paper are from the PS1-MDS. We refer the reader to Chambers et al (2016) for details of the PS1 survey telescope and PS1-MDS observing strategy, and to Villar et al (2020) and Hosseinzadeh et al (2020) for the definition of the overall sample of SN-like transients and their light curves, description of the sub-sample of spectroscopically-classified events, the photometric classification approaches and results, all relevant data (including photometry and host galaxy redshifts), and complete descriptions of the algorithms and training processes.…”

Section: Sample Constructionmentioning

confidence: 99%

“…In this paper we focus on the sample of photometrically-classified SLSNe. 1 Using SuperRAENN (Villar et al 2020) and Superphot (Villar et al 2018;Hosseinzadeh et al 2020) we photometrically classified 58 and 37 SLSNe, respectively, using the same training set of 557 spectroscopically-classified SNe, which includes 17 SLSNe that were studied in Lunnan et al (2018). Here, we adopt the class with the highest probability as the predicted SN type for each transient.…”

Section: Sample Constructionmentioning

confidence: 99%

See 1 more Smart Citation

Photometrically-Classified Superluminous Supernovae from the Pan-STARRS1 Medium Deep Survey: A Case Study for Science with Machine Learning-Based Classification

Hsu,

Hosseinzadeh,

Villar

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

With the upcoming Vera C. Rubin Observatory Legacy Survey of Space and Time (LSST), it is expected that only ∼ 0.1% of all transients will be classified spectroscopically. To conduct studies of rare transients, such as Type I superluminous supernovae (SLSNe), we must instead rely on photometric classification. In this vein, here we carry out a pilot study of SLSNe from the Pan-STARRS1 Medium-Deep Survey (PS1-MDS) classified photometrically with our SuperRAENN and Superphot algorithms. We first construct a sub-sample of the photometric sample using a list of simple selection metrics designed to minimize contamination and ensure sufficient data quality for modeling. We then fit the multi-band light curves with a magnetar spin-down model using the Modular Open-Source Fitter for Transients (MOSFiT). Comparing the magnetar engine and ejecta parameter distributions of the photometric sample to those of the PS1-MDS spectroscopic sample and a larger literature spectroscopic sample, we find that these samples are overall consistent, but that the photometric sample extends to slower spins and lower ejecta masses, which correspond to lower luminosity events, as expected for photometric selection. While our PS1-MDS photometric sample is still smaller than the overall SLSN spectroscopic sample, our methodology paves the way to an orders-of-magnitude increase in the SLSN sample in the LSST era through photometric selection and study.

show abstract

“…Machine learning techniques will be required to separate SLSNe from other transients in this enormous data stream (e.g. Gomez et al 2020;Villar et al 2020;Hosseinzadeh et al 2020;Muthukrishna et al 2019), and enable spectroscopic and multi-wavelength follow-up. With thousands of SLSNe, we will much more finely sample their diversity and energetics, and continue to pin down their progenitors.…”

Section: Looking Aheadmentioning

confidence: 99%

Superluminous supernovae: an explosive decade

Nicholl

2021

Preprint

View full text Add to dashboard Cite

I review our current understanding of superluminous supernovae, mysterious events 100 times brighter than conventional stellar explosions. INTRODUCTION: THE DISCOVERY OF SLSNEThe study of supernovae spans centuries and cultures, with Chinese astronomers recording the oldest known 'guest star' in 185 AD. By the turn of the millenium, it was well established that core-collapse supernovae (CCSNe) signal the death of a star of more than 8 M (where M is the solar mass). When the nuclear reactor in its core has converted all its fuel into stable iron, and can no longer extract energy to support itself, it collapses to a neutron star and releases its gravitational energy. Type Ia supernovae (SNe Ia) occur when a white dwarf gains sufficient mass from a binary companion to encounter a runaway thermonuclear reaction. Both types of supernova release ∼ 10 51 erg in kinetic energy, peak with a luminosity up to 10 9 L , and produce many of the chemical elements needed for life in the Universe.A more extreme class of 'superluminous supernovae' (SLSNe), ∼ 10 − 100 times brighter again, escaped our notice until the 21st century. Their eventual discovery was intimately tied with advances in optical sky surveys in the mid-to-late 2000s. Figure 1 shows the growth in the supernova discovery rate from 1996-2020, with a sharp uptick around 2010 driven by wide-field robotic telescopes and automated source detection employed by surveys such as the Catalina Real-time Transient Survey (CRTS; Drake et al. 2009), the Panoramic Survey Telescope and Rapid Reponse System (PanSTARRS; Kaiser et al. 2010) and the Palomar Transient Factory (PTF; Rau et al. 2009).Nearly 20,000 supernovae were reported in 2020 -a 100-fold increase in just 20 years. The volumetric rate of SLSNe is only ∼1 in every few thousand supernovae (Quimby et al. 2013;Frohmaier et al. 2021), though because these brighter explosions can be detected out to greater distances, they should make up ∼1% of the supernovae detected by a given survey. Finding SLSNe in significant numbers therefore requires the discovery and classification of hundreds or thousands of supernovae per year.Yet Figure 1 also shows that the fraction (not just the absolute number) of SLSNe within a given survey also increased dramatically in 2010. Earlier surveys had targeted massive galaxies that have a correspondingly high rate of CCSNe and SNe Ia, but it turns out that SLSNe

show abstract

Photometric Classification of 2315 Pan-STARRS1 Supernovae with Superphot

Cited by 25 publications

References 110 publications

ParSNIP: Generative Models of Transient Light Curves with Physics-Enabled Deep Learning

ParSNIP: Generative Models of Transient Light Curves with Physics-Enabled Deep Learning

Photometrically-Classified Superluminous Supernovae from the Pan-STARRS1 Medium Deep Survey: A Case Study for Science with Machine Learning-Based Classification

Superluminous supernovae: an explosive decade

Contact Info

Product

Resources

About