|
The deferred photometric pipeline of SNLS uses only photometry to detect and classify type Ia supernovae (SNe Ia). The advantages of a photometric sample include larger number of events classified as type Ia, larger redshift coverage and no need for spectroscopic observations. In this seminar I will present improvements on the detection of transient events for this pipeline and SNe classification using photometric redshifts obtained directly from light curves. First, a subtracted image stack treatment was developed to reduce the number of non SN-like events using morphological component analysis. This technique exploits the morphological diversity of objects to be detected to extract the signal of interest. With this method spurious detections were reduced and coordinate resolution was improved for the SNLS 3 year sample. Second, a new classification using photometric SN redshifts is introduced. This algorithm provides redshifts for all events with a better average precision and lower catastrophic errors than the host galaxy photometric redshift catalogue used in the previous analysis. The classification was optimised by machine learning algorithms which provide a sample of SN Ia with estimated high purity. |
|