COSMOLOGY WITH PHOTOMETRICALLY CLASSIFIED TYPE Ia SUPERNOVAE FROM THE SDSS-II SUPERNOVA SURVEY

We present the cosmological analysis of 752 photometrically–classified Type Ia Supernovae (SNe Ia)
\nobtained from the full Sloan Digital Sky Survey II (SDSS-II) Supernova (SN) Survey, supplemented
\nwith host–galaxy spectroscopy from the SDSS-III Baryon Oscillation Spectroscopic Survey (BOSS).
\nOur photometric–classificationmethod is based on the SN typing technique of Sako et al. (2011), aided
\nby host galaxy redshifts (0.05 < z < 0.55). SNANA simulations of our methodology estimate that
\nwe have a SN Ia typing efficiency of 70.8%, with only 3.9% contamination from core-collapse (non-Ia)
\nSNe. We demonstrate that this level of contamination has no effect on our cosmological constraints.
\nWe quantify and correct for our selection effects (e.g., Malmquist bias) using simulations. When fitting
\nto a flat _CDM cosmological model, we find that our photometric sample alone gives Ωm = 0.24+0.07/−0.05
\n(statistical errors only). If we relax the constraint on flatness, then our sample provides competitive
\njoint statistical constraints on Ωm and Ω∆, comparable to those derived from the spectroscopically-
\nconfirmed three-year Supernova Legacy Survey (SNLS3). Using only our data, the statistics–only
\nresult favors an accelerating universe at 99.96% confidence. Assuming a constant wCDM cosmological
\nmodel, and combining with H0, CMB and LRG data, we obtain w = −0.96+0.10/−0.10, Ωm = 0.29+0.02/−0.02 and Ωk = 0.00+0.03/−0.02 (statistical errors only), which is competitive with similar spectroscopically confirmed
\nSNe Ia analyses. Overall this comparison is reassuring, considering the lower redshift leverage of the
\nSDSS-II SN sample (z < 0.55) and the lack of spectroscopic confirmation used herein. These results
\ndemonstrate the potential of photometrically–classified SNe Ia samples in improving cosmological
\nconstraints.