Supernovae from LSST will test our models of the Universe

Supernovae, the most massive explosions in the Universe, are powerful probes of cosmology. LSST will discover an unprecedented number of supernovae, very few of which will be followed up to have their types confirmed. Because only one type of supernova is useful for cosmology, classification from a supernova's light curve (how its brightness decays with time) is essential. While light curves from different supernova types are difficult to distinguish by eye (Figure 1), we can use mathematical techniques to decompose the light curve into features that are easy to separate (Figure 2). From these features, we are able to use machine learning to train a computer to automatically classify large numbers of supernovae. 

peiris_fig_1_light_curve.png

Figure 1: Light Curve.

Figure 2: Classification of features.

Credit: Michelle Lochner, Hiranya Peiris, Jason McEwen, Ofer Lahav/ UCL

Last updated: 14 Dec 2015 at 14:30