Cataclysmic Variable Classifier

Cataclysmic Variable stars (CVs) are binary star systems where a white dwarf (WD) accretes material off of what is usually a low-mass, main-sequence star. Depending on how strong the magnetic field strength of the white dwarf is, this material might be channelled down onto the surface of the white dwarf via magnetic field lines, or it might spiral down onto the surface and form a disk of accreting material.

Image credit: : K. Smale

CVs can be broadly distributed in the following categories:

  • Classical novae: These have only one observed eruption in which hydrogen-rich material accretes on to the surface of the WD.
  • Dwarf Novae: Where the release of gravitational energy caused by a temporary large increase in the rate of mass transfer through the accretion disk.
  • Recurrent novae: Much like classical novae, but repeat their eruptions.
  • Nova-like variables: All ‘non-eruptive’ CVs.
  • Magnetic CVs: Where magnetic fields of the primaries can disrupt the accretion disk partially or totally.

This project consists of two parts which aims to help classify CVs and in which category of CV they belong. The first part of the project is to design a photometric “field guide” for citizen scientists to identify CVs by their light curve. The second, and most challenging, part of the project aims to design a machine learning classifier that would recognize whether or not a transient is a CV soon after it is first detected. This classifier would be useful for projects like the Young Supernova Experiment where objects of interest are often mistaken for CVs.