Caltech astronomers have developed a machine learning algorithm that successfully classified 1,000 supernovas caused by dying stars as they explode. The algorithm is called SNIascore and it uses applied data captured by the Zwicky Transient Facility (ZTF), a sky survey instrument at the Palomar Observatory.
“We needed a helping hand, and we knew that they would take a big load off our backs once we trained our computers to do the job. SNIascore classified its first supernova in April 2021, and, a year and a half later, we are hitting a nice milestone of 1,000 supernovae,” says Christoffer Fremling, a staff astronomer at Caltech, in a press statement. Fremling is the mastermind behind SNIascore
Tons of data
Every night, ZTF scans the sky and looks for changes called transient events. These can include anything from moving asteroids to black holes that have just eaten stars or supernovae. Throughout the night, ZTF sends hundreds of thousands of alerts to astronomers worldwide, notifying them of such transient events.
Based on these alerts, astronomers can use other telescopes to follow up and investigate the event. But the sheer volume of the data being collected by the facility every night means that ZTF team members cannot sort through the data on their own. This is where SNIAscore comes in.
“The traditional notion of an astronomer sitting at the observatory and sieving through telescope images carries a lot of romanticism but is drifting away from reality,” says Matthew Graham, project scientist for ZTF, in a press statement. Graham is also a professor of astronomy at Caltech.
SNIascore works with this machine to classify supernovae that are Type Ia. When it does this, it provides a reliable stream of data on supernovae that scientists can further investigate. According to Caltech, SNIascore is “remarkably accurate” and hasn’t misclassified a single one of the 1,000 supernovae.