Astronomers discover 116,027 new variable stars

Using machine learning techniques and data from the All-Sky Automated Survey for Supernovae (ASAS-SN) and several other surveys, astronomers have identified 378,861 variable stars, of which 262,834 are known variables and 116 027 are new discoveries.

Variable stars are celestial objects that wax and wane in brightness over time, especially when viewed from our vantage point on Earth.

“In fact, even our Sun is considered a variable star,” said Dr. Collin Christy, an astronomer in the Department of Astronomy at Ohio State University.

“Surveys like ASAS-SN are a particularly important tool for finding systems that can reveal the complexity of stellar processes.”

“Variable stars are a bit like a stellar laboratory. These are really cool places in the universe where we can study and learn about how stars actually work and the little intricacies they all have.

In the study, Dr. Christy and his colleagues analyzed data from ASAS-SN as well as ESA’s Gaia mission, the Two Micron All Sky Survey (2MASS) and the AllWISE catalog.

They used a machine learning algorithm to generate a list of 1.5 million candidate variable stars from a catalog of around 55 million isolated stars.

Then they further reduced the number of applicants. Of the 1.5 million stars they studied, 378,861 turned out to be true variable stars.

More than half were already known to the astronomical community, but 116,027 of them turned out to be new discoveries, including more than 111,000 periodic variables and 5,000 irregular variables.

“We plan to integrate these variables, including low-probability candidates, into our Citizen Science initiative to help refine our classifications and improve our machine learning techniques,” the astronomers said.

“Citizen scientists have outperformed our current machine learning classifier in identifying parasitic variables.”

“They also excelled in identifying unusual or highly variable candidates.”

“This is the first time that we really combine citizen science with machine learning techniques in the field of variable star astronomy,” said Tharindu Jayasinghe, a doctoral student in the Department of Astronomy and the Center for Cosmology and of Astroparticle Physics from Ohio. State University.

“We’re expanding the boundaries of what you can do when you combine these two things.”

The results will be published in the Royal Astronomical Society Monthly Notices.

_____

CT Christy et al. 2022. The ASAS-SN Catalog of X Variable Stars: Discovery of 116,000 New Variable Stars Using g-Band Photometry. MNRAS, in the press; arXiv: 2205.02239