For the first time in the world, artificial intelligence discovered a supernova without human help
[ad_1]
For the first time, artificial intelligence found, detected, confirmed, classified and announced the discovery of a supernova without human intervention.
It was named SN2023tyk, according to Northwestern University in Illinois, USA.
On the basis of this educational institution, an international team of scientists worked, who developed Bright Transient Survey Bot (BTSbot) – a new tool of artificial intelligence. They used more than 1.4 million images from nearly 16,000 sources to train an AI algorithm.
“Ultimately, taking humans out of the loop gives the research team more time to analyze their observations and develop new hypotheses to explain the origin of the cosmic explosions we observe.”said lead researcher and astronomer Adam Miller.
| Photo: NASA |
According to his colleague, Nabil Rehemtullah, this will greatly facilitate large-scale studies of supernovae (a type of star that explodes in binary systems, where one of the stars is a white dwarf and the other may be a giant or another type of star) and help to better understand their life cycles. as well as the origin of the elements from which they are made.
The supernova, called SN2023tyk, was discovered by AI in data from the California Robotics Camera (ZTF), which scans the northern sky every two days.
According to scientists, the explosion in space occurred on October 3, 2 days later BTSbot recorded it, and on October 7 – classified it and “announced” the discovery of a supernova.
Astronomers note that despite the fact that such objects are bright, they are quite difficult to notice. Traditional detection methods rely on astronomers visually inspecting large volumes of data from robotic telescopes that constantly scan the night sky for new sources of light.
“The ZTF has been operating for the past 6 years, during which time I and other scientists have spent more than 2,000 hours visually inspecting star candidates and determining which ones to observe using spectroscopy.
Adding BTSbot to our workflow will save us from having to spend so much time checking“, says astronomer Christopher Framling of the California Institute of Technology.
Scientist Miller notes that automated software will provide researchers with a list of potential explosions.
“We can definitively know that the candidate is indeed a supernova by collecting it the spectrum is the scattered light of the source, which reveals the elements present in the explosion“he added.
This is a time-consuming process, and astronomers estimate that only a small fraction of all supernovae that occur in the universe have been detected.
BTSbot automatically “reached out” to another robotic instrument called the Spectral Energy Distribution Machine (SEDM) to conduct a broad observation of the potential supernova to collect its spectrum.
After receiving it, SEDM sent it to Caltech’s SNIascore for supernova classification.
White dwarf supernovae are especially important to astronomers because they can be used to measure the expansion of the universe.
We will remind you that in April, scientists discovered supernova stars that can destroy Earth-like planets with their explosion.
Read also: The Hubble telescope photographed a distant galaxy in incredible colors. PHOTO
[ad_2]
Original Source Link
