Off-line processing and identification of SNLS spectra using PHASE (PHotometry Assisted Spectra Extraction) and SALT2
S. Baumont, C. Balland & SNLS collaboration

Abstract: In this poster, we present the first results of the off-line processing of, SNLS spectra using a dedicated pipeline that aims at: 1) producing an homogeneous set of supernova and host spectra whose extraction has been optimized using all available information, including photometric priors from CFHT images to derive the, spectral trace of the SN; 2) identifying supernovae using both light-curves, and the spectral template of SALT2 (Guy etal 2006). Resampling of data is avoided so as to preserve the independance of pixels. This allows a robust estimation and propagation of the statistical errors. We find that PHASE extractions are often improved over real-time analysis in terms of signal-to-noise ratio and host residual contamination, except when the supernova is unresolved from the host (high redshift, and-or small host). In this case, the two extractions are similar. SALT2 then provides a clean identification of these spectra in most cases. These results are used to validate the supernovae entering the Hubble diagram.