AstroNote 2020-195

Primary tabs

DRAFT
2020-10-16 19:12:18
Type: Object/s-Discovery/Classification
Spectroscopic Classification of Four Superluminous Supernovae Using FLEET
Authors: Sebastian Gomez (Harvard), Edo Berger (Harvard), Peter K. Blanchard (Northwestern), Griffin Hosseinzadeh (Harvard), Matt Nicholl (Birmingham), V. Ashley Villar (Columbia), Yao Yin (Harvard)
Source Group: FLEET
Abstract:
We report on the spectroscopic classification of three Type-I superluminous supernovae (SLSNe) and one Type-II SLSN, in addition to a Type-Ia and a Type-II SN. These SNe were selected with the Finding Luminous and Exotic Extragalactic Transients (FLEET) pipeline as likely SLSN candidates (Gomez et al. 2020).

The Finding Luminous and Exotic Extragalactic Transients (FLEET) pipeline is a machine learning pipeline designed to predict the probability, P(SLSN), of any new transient being a SLSN.  FLEET uses a random forest algorithm trained on every classified transient from the TNS to derive a P(SLSN) value. The features used to separate SLSNe from other transients include light curve information, as well as contextual (host galaxy) information.  We find that maximum classification accuracy can be achieved for transients with P(SLSN) > 0.8, where we expect ~85% of them to be SLSNe; for events with P(SLSN) > 0.5 we expect ~50% of them to be SLSNe. FLEET is described in detail in Gomez et al. 2020 (ApJ in press; arXiv:2009.01853).

Here we present the first results from FLEET.  We used the pipeline to generate a list of SLSN candidates, and obtained optical spectroscopy of six SLSN candidates using the MMT 6.5-m telescope. We classified the spectra using template matching with SNID (Blondin & Tonry, 2007, ApJ, 666, 1024). We find the classifications to be in agreement with the predictions from our random forest classifier: The 2 objects with P(SLSN) > 0.8 are both classified as SLSNe, and 4 of the 6 of objects with P(SLSN) > 0.5 are classified as SLSNe.

Gomez, S. Berger, E., Blanchard, P. K., et al. FLEET: A Redshift-Agnostic Machine Learning Pipeline to Rapidly Identify Hydrogen-Poor Superluminous Supernovae. 2020, Accepted to ApJ. https://ui.adsabs.harvard.edu/abs/2020arXiv200901853G

Show current TNS values
CatalogNameReported RAReported DECReported Obj-TypeReported RedshiftHost NameHost RedshiftP(SLSN)Obs. DatePeak abs. mag.RemarksTNS RATNS DECTNS Obj-TypeTNS Redshift
TNS2020jii15:34:55.291+02:51:11.66SLSN-I0.3960.912020-07-15-22.215:34:55.291+02:51:11.66SLSN-I0.396
TNS2020myh23:37:30.549+21:42:42.81SLSN-I0.2830.812020-09-15-21.523:37:30.549+21:42:42.81SLSN-I0.283
TNS2020hvw17:58:29.929+23:56:21.27SN II0.0930.702020-09-17-19.117:58:29.929+23:56:21.27SN II0.093
TNS2020mad14:15:52.740+05:26:35.63SLSN-II0.1230.662020-07-19-20.114:15:52.740+05:26:35.63SLSN-II0.123
TNS2020oqy20:37:45.829+69:55:08.52SN Ia0.1350.642020-08-19-19.320:37:45.829+69:55:08.52SN Ia0.135
TNS2020onb14:23:00.600+49:10:40.69SLSN-I0.1530.492020-07-17-20.114:23:00.600+49:10:40.69SLSN-I0.153