By tracking atypical spectral distributions in 12 filters of the J-PLUS survey, astronomers isolate hundreds of rare candidates to be confirmed by spectroscopy.
Finding rare objects in a sky survey before even having their spectrum is a major challenge in preparing for monitoring campaigns. A team applied an autoencoder-based anomaly detection method to photometric data from the Javalambre Photometric Local Universe Survey (J-PLUS), whose fourth data release covers twelve filters, both wide and narrow. The tool reports spectral energy distributions that deviate from the dominant sample behavior.
Islands of special populations
The atypical objects were then projected with the t-SNE algorithm according to their colors, then cross-referenced with SIMBAD and other surveys to identify the clusters. These islands correspond to well-defined populations: many white dwarfs and low-mass stars, including a small group of white dwarfs "polluted" by heavy elements.
Above all, the study draws up a list of targets previously absent from the literature: 19 carbon-rich stars, 11 cataclysmic variables and 170 binary systems associating a white dwarf with a main sequence star, proposed for spectroscopic monitoring.
Read the full article on the original source
Go to the source →