Rewind supernovae with machine learning

Supernovae are catastrophic stellar explosions shaping the visible Universe. They play an important role in the synthesis and distribution of almost all elements and especially heavy elements such as iron, enriching the Universe since the first supernova explosion when the cosmos was metal-free. We are now in the golden era of supernova astronomy - and in general of transients - since astronomical surveys are discovering roughly 11000 transients per year. Unfortunately, the majority of them remain unrecognised, severely diminishing the scientific return. Future astronomical experiments (for example the Large Synoptic Survey Telescope - LSST) will make it more challenging boosting the number of yearly discoveries by a factor of 100, making almost impossible to recognise the different types of supernovae in the first days of their evolution, when important information about the nature of the object can be obtained.

Nevertheless, it is usually easier and possible to gather information at a later time (e.g. more than 50 days after the explosion of the star). However, we are still missing the tools to connect the early time (less than 30 days from the explosion) information of a supernova, such as luminosity evolution and electromagnetic spectra signatures, with the later ones. To connect late to early time data/observations several approaches are possible. Among them, the application of supervised and unsupervised learning algorithms to large datasets, supplied by surveys such as the Public ESO Spectroscopic Survey of Transient Objects (PESSTO) of which the supervisor is a member, provides an innovative view of the issue. This methodology will give the possibility to recognise the transient but also to retrieve the information carried by the early electromagnetic behaviour (even when not available) and shed lights on the supernova physics and its connection to the environment in which they explode.

In this project, the PhD student will gather knowledge of supernova explosions linked to the life and death of massive stars as well as programming skills in python and experience in observational astronomy, data reduction and data analysis.