Latest News And Breaking Headlines

AI helps discover new space anomalies


Credit: Pixabay/CC0 public domain

The SNAD team, an international network of researchers including Matvey Kornilov, associate professor of the HSE University Faculty of Physics, has discovered 11 previously undiscovered space anomalies, seven of which are supernova candidates. The researchers analyzed digital images of the northern sky taken in 2018 using a kD tree to detect anomalies via the “nearest neighbor” method. Machine learning algorithms helped automate the search. The newspaper is published in New astronomy.

Most astronomical discoveries are based on observations with subsequent calculations. Although the total number of observations in the 20th century was still relatively small, the amounts of data increased dramatically with the advent of large-scale astronomical studies. For example, the Zwicky Transient Facility (ZTF), which uses a wide-angle camera to survey the northern sky, generates ∼1.4 TB of data per night of observation, and its catalog contains billions of objects. Handling such massive amounts of data is both expensive and time-consuming, so the SNAD team of researchers from Russia, France and the US came together to develop an automated solution.

When scientists examine astronomical objects, they observe their light curves, which show variations in an object’s brightness as a function of time. The observers first identify a flash of light in the sky and then follow its evolution to see if the light brightens, dims or goes out over time. In this study, the researchers examined one million real light curves from the 2018 ZTF catalog and seven simulated live curve models of the types of objects being studied. In total, they tracked about 40 parameters, including the amplitude of an object’s brightness and the time frame.

“We described the properties of our simulations using a set of features that are expected to be observed in real astronomical bodies. In the dataset of approximately one million objects, we were looking for super-powerful supernovae, Type Ia supernovae, Type II supernovae, supernovas and tidal disturbances,” explains Konstantin Malanchev, co-author of the paper and postdoctoral fellow at the University of Illinois at Urbana-Champaign. “We refer to such classes of objects as anomalies. They are either very rare, with little known properties, or seem interesting enough to merit further investigation.”

The light curve data from real objects were then compared to that from simulations using the kD tree algorithm. A kD tree is a geometric data structure for dividing space into smaller parts by cutting them with hyperplanes, planes, lines or points. In the current study, this algorithm was used to narrow the search range when looking for real objects with properties similar to those described in the seven simulations.

Next, the team identified 15 closest neighbors, i.e., real objects from the ZTF database, for each simulation — 105 matches in all, which the researchers then visually examined to check for anomalies. The manual verification confirmed 11 anomalies, including seven supernova candidates and four active galactic core candidates where tidal disturbances could occur.

“This is a very good result,” said Maria Pruzhinskaya, co-author of the paper and research fellow at the Sternberg Astronomical Institute. “In addition to the rare objects already discovered, we were able to detect several new ones that were previously missed by astronomers. This means that existing search algorithms can be improved to avoid missing such objects.”

This research shows that the method is very effective, but relatively easy to apply. The proposed algorithm for detecting space phenomena of a certain type is universal and can be used to discover all interesting astronomical objects, not limited to rare types of supernovae.

“Astronomical and astrophysical phenomena that have not yet been discovered are in fact anomalies,” said Matvey Kornilov, associate professor of the HSE University Faculty of Physics. “Their observed manifestations are expected to differ from the properties of known objects. In the future, we will try to use our method to discover new classes of objects.”

A new anomaly detection pipeline for astronomical discovery and recommendation systems

More information:
PD Aleo et al, SNAD transient miner: finding missed transients in ZTF DR4 using kD trees, New astronomy (2022). DOI: 10.116/j.nevast.2022.101846

Provided by National Research University Higher School of Economics

Quote: AI helps discover new space anomalies (2022, August 5) retrieved August 6, 2022 from https://phys.org/news/2022-08-ai-space-anomalies.html

This document is copyrighted. Other than fair dealing for personal study or research, nothing may be reproduced without written permission. The content is provided for informational purposes only.

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept Read More