Bayesian Single-Epoch Photometric Classification of Supernovae

Abstract

Ongoing supernova (SN) surveys find hundreds of candidates that require confirmation for their various uses. Traditional classification based on follow- up spectroscopy of all candidates is virtually impossible for these large samples. The use of Type Ia SNe as standard candles is at an evolved stage that requires pure, uncontaminated samples. However, other SN survey applications, such as measuring cosmic SN rates, could benefit froma classification of SNe on a statistical basis, rather than case by case. With this objective in mind, we have developed the SN-ABC, an automatic Bayesian classifying algorithm for supernovae. We rely solely on single- epoch multiband photometry and host-galaxy (photometric) redshift information to sort SN candidates into the two major types, Ia and core-collapse supernovae. We test the SN-ABC performance on published samples of SNe from the Supernova Legacy Survey (SNLS) and GOODS projects that have both broadband photometry and spectroscopic classification (so the true type is known). The SN- ABC correctly classifies up to 97% (85%) of the Type Ia (II-P) SNe in SNLS, and similar fractions of the GOODS SNe, depending on photometric redshift quality. Using simulations with large artificial samples, we find similarly high success fractions for Types Ia and II-P, and reasonable (~75%) success rates in classifying Type Ibc SNe as core-collapse. Type IIn SNe, however, are often misclassified as Type Ia. In deep surveys, SNe Ia are best classified at redshifts z ≳ 0.6 or when near maximum. Core-collapse SNe (other than Type IIn) are best recognized several weeks after maximum, or at z ≾ 0.6. Assuming the SNe are young, as would be the case for rolling surveys, the success fractions improve by a degree dependent on the type and redshift. The fractional contamination of a single-epoch photometrically selected sample of SNe la by core-collapse SNe varies between less than 10% and as much as 30%, depending on the intrinsic fraction and redshift distribution of the core-collapse SNe in a given survey. The SN-ABC also allows the rejection of SN impostors such as active galactic nuclei (AGNs), with half of the AGNs we simulate rejected by the algorithm. Our algorithm also supplies a good measure of the quality of the classification, which is valuable for error estimation.