Astronomers have spotted more than 27,000 killer asteroids, some of which are passing dangerously close to Earth

Maria ShevchukNews
Detection of dangerous asteroids will help to create a "global defense" in time. Source: Pixabay

Astronomers and data scientists working on powerful computers from Google have discovered 27,500 new asteroids. Some of them are passing dangerously close to Earth.

The software for hunting for space bodies will contribute to the development of a "comprehensive map of the solar system," the Daily Mail writes. According to one of the developers, Harvard astrophysicist Matthew Holman, this will help protect the planet from dangerous bodies.

Almost 1.7 billion points of light are documented in 412,000 infrared images from the archives of the US National Optical Infrared Astronomy Laboratory (NOIRLab). These photos were scanned using a new algorithm for detecting hazardous asteroids developed as part of the project.

"This is the key to protecting the Earth from asteroid impacts, namely knowing where they are," said Ed Lu, one of the former NASA astronauts who led the project.

Since 2002, Ed Lu has been the executive director of the nonprofit organization B612, where he heads the group's Asteroid Institute. The Institute is currently playing a leading role in the discovery of new objects of cosmic origin.

"The interesting thing is that we don't have our own telescope. We do it with data science," Ed said.

The Asteroid Institute's B612 algorithm, known as Tracklet-less Heliocentric Orbit Recovery (THOR), analyzed NOIRLab's archive of astronomical images to identify space objects that threaten our planet.

Based on data from just two images, sometimes taken on different nights and with two different telescopes, THOR was able to calculate whether these points of light were actually one asteroid in orbit around our solar system.

Astronomers have spotted more than 27,000 killer asteroids, some of which are passing dangerously close to Earth

"We are using electrons in data centers in addition to conventional photons in telescopes," Dr. Lu said.

The THOR algorithm was developed by the B612 Asteroid Institute in collaboration with the University of Washington's DiRAC Institute.

The limited computing power required to detect viable asteroids in these scattered telescope images would have made this process nearly impossible at one time, even in the recent past.

But Google Cloud's distributed network of computing power allowed THOR to identify 27,500 new "asteroid candidates" in just five weeks.

According to a press statement from the nonprofit organization B612, their collaboration with Google Cloud will eventually allow THOR to process 5.4 billion observations from various astronomical observatories and equipment. These are celestial mysteries that could turn out to be asteroids or other space objects.

The Asteroid Institute has announced that it is also exploring the possibility of using Google's artificial intelligence technology to automate the screening and verification of potential asteroids, as first reported by their THOR algorithm.

In the past, the initial screening of these candidates has relied on an army of volunteers: high school students, undergraduates, graduate students, scientists, and professional astronomers.

But, according to Dr. Lu, if the project's expansion with artificial intelligence is successful, reducing this time-consuming verification work will help the institute adapt the THOR process for much larger astronomical datasets.

Only verified information is available on OBOZ.UA Telegram channel and Viber. Do not fall for fakes!

Other News

How to make mayonnaise without oil and eggs: two options

How to make mayonnaise without oil and eggs: two options

For salads, marinades, baking