Asteroid hunters discover more than 27,000 ‘hidden’ space rocks – and some pass ‘dangerously’ close to Earth

Astronomers and data scientists working with a boost of computing power borrowed from Google have discovered 27,500 new asteroids, some of them dangerously close to Earth.

Their collaboration hopes to accelerate the development of “a comprehensive map of the solar system” needed for “planetary defense,” according to a Harvard astrophysicist, Matthew Holman, who helped develop the asteroid-hunting software.

Nearly 1.7 billion points of light, documented in 412,000 infrared images from the archives of the US National Optical-Infrared Astronomy Research Laboratory (NOIRLAb), were scanned by the new asteroid hunting algorithm.

“This is super important,” as a former NASA astronaut who led the project said.

“This is the key to protecting Earth from asteroid attack: knowing where they all are.”

As a former NASA astronaut who led the project put it, “This is the key to protecting Earth from asteroid collisions: knowing where they all are.”

Since 2002, that retired NASA astronaut, a Stanford-educated astrophysicist Dr. Ed Luwas executive director of the nonprofit B612, where he now heads the group’s Asteroid Institute.

“Right now, we are now one of the leading discoverers of asteroids on Earth,” said Dr. Lu, as he discussed the team’s identification of tens of thousands of new asteroids.

‘But what makes this interesting is that we don’t own a telescope. We don’t operate a telescope,” he continued. ‘We do this from a data science perspective.’

The B612 Asteroid Institute’s algorithm – known as the Tracklet-less Heliocentric Orbit Recovery, or THOR – searched NOIRLAb’s astronomical image archive to distinguish which reflected points of light were actually nearby asteroids.

Using evidence from just two images, sometimes taken on different nights and using two different telescopes, THOR could calculate whether those points of light were in fact one asteroid orbiting our solar system.

“The exciting thing is that we are using electrons in data centers, in addition to the usual photons in telescopes, to make astronomical discoveries,” said Dr. Lu.

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

Since 2002, retired NASA astronaut and Stanford-educated astrophysicist Dr. Ed Lu is executive director of the nonprofit B612, where he now heads the group’s Asteroid Institute. Top right a NASA photo of Dr. Lu, taken during his astronaut career on October 10, 2000

Nearly 1.7 billion points of light documented in 412,000 images from the digital archives of the US National Optical-Infrared Astronomy Research Laboratory (NOIRLab) were scanned by the project’s new asteroid hunting software. Above is another view of the new asteroid map

The raw computing power required to identify viable asteroid candidates in these disparate telescope images would once have made this process virtually impossible, even in the recent past.

But thanks to Google Cloud’s distributed network of computing power, THOR was able to identify its 27,500 new, “highly confident asteroid discovery candidates” in just about five weeks.

“This is an example of what’s possible,” said Massimo Mascaro, technical director of the Office of the Chief Technology Officer of Google Cloud.

“I can’t even quantify how many opportunities there are in terms of data already collected there, and if analyzed with the right calculation, they could lead to even more results,” Mascaro explains.

According to a press statement from the nonprofit B612, their partnership with Google Cloud will ultimately put THOR to work on 5.4 billion observations from various astronomical surveys and equipment – ​​celestial mysteries that may turn out to be asteroids or other objects in space.

The Asteroid Institute announced that it is also exploring the feasibility of using Google’s AI technologies in an effort to automate the vetting and verification of potential asteroid candidates, as first reported by their THOR algorithm.

In the past, the initial verification of these candidates was left to a volunteer army of high school students, college students, postdoctoral researchers, scientists and professional astronomers.

But if these AI-enabled extensions of their project are successful, said Dr. Lu said the reduction in this time-consuming human verification work will help the institute adapt THOR’s process for much more extensive astronomical data sets, such as those that will soon be available. the Vera C. Rubin Observatory, now under construction in Chile.

“This is a big change,” said Dr. Lu.