Astronomie - Has LIGO Seen Galaxy-Warped Gravitational Waves?


Nobel laureate George Smoot claims LIGO has observed amplified signals of black hole mergers from the very distant universe, but LIGO scientists disagree


A visualization of all 10 black-hole mergers (and their gravitational-wave signals) announced by the LIGO collaboration. According to a new study, two mergers depicted in the bottom row — GW170809 and GW170814 — may actually be twin images of a single event, split and magnified by the gravitational effects of an intervening galaxy. Credit: Teresita Ramirez/Geoffrey Lovelace/SXS Collaboration/LIGO-Virgo Collaboration


Announced by the LIGO collaboration in February 2016, the discovery of ripples in spacetime known as gravitational waves was momentous enough to merit the 2017 Nobel Prize in Physics. Now, another Nobel laureate says LIGO has unknowingly made another spectacular discovery: gravitational waves from merging black holes that have been amplified by the gravity of intervening galaxies.

Called gravitational lensing, this phenomenon is routinely used to study light from objects in the very distant cosmos. But the new assertion, if proved correct, would make it the first such sighting for gravitational waves. The controversial claim, which has been dismissed by members of the LIGO team, comes via physics Nobelist George Smoot of the University of California, Berkeley, and his colleagues. “We are wagering our reputations on this,” he says.

LIGO (for the Laser Interferometer Gravitational-Wave Observatory), comprising two detectors in the U.S., and Virgo, a detector outside Pisa, Italy, have together so far announced observations of gravitational waves from the merging of 10 pairs of black holes as well as a pair of neutron stars.

It is the black holes that Smoot has in his sights. According to a LIGO–Virgo analysis, the black hole mergers occurred in the relatively nearby universe, generally a few billion light-years from Earth. Many of the black holes that merged had a mass of about 30 suns, and one was about 50 solar masses. Such black holes are formed by the gravitational collapse of giant stars. According to Smoot, our galaxy and by extension the nearby universe surveyed by LIGO is lacking in the sort of stars—known as low-metallicity stars—required to form such big black holes. If so, this should lead to a local paucity of black holes with the range of masses seen by LIGO. “The maximum you can make is about 20 solar masses,” Smoot says. “Except in very unusual situations.”

In support of their argument, Smoot and astrophysicists Tom Broadhurst of the University of Basque Country in Bilbao, Spain, and José María Diego of the University of Cantabria in Santander, Spain, point to x-ray surveys of black hole pairs in the Milky Way that suggest the mass distribution of such black holes peaks around 10 solar masses. Presuming this same distribution holds for black holes in the greater volume of space studied by LIGO, Smoot and colleagues argue the higher black hole mass estimates from the LIGO–Virgo team must be a miscalculation. Rather than seeing gravitational waves from abnormally large merging black holes in the nearby universe, Smoot and his colleagues say, LIGO and Virgo are actually seeing smaller merger events taking place much farther away—on the order of 10 billion light-years distant—magnified and made visible through gravitational lensing.


According to Einstein’s general relativity, gravitational lenses form because galaxies and galaxy clusters noticeably warp spacetime. If a galaxy lies between Earth and some distant object, then that galaxy behaves like a lens, curving spacetime to magnify that object’s light as seen from Earth. Gravitational waves also must follow curved spacetime—so they too can be lensed and magnified by gravitational lenses. Furthermore, the greater an object’s distance from Earth, the greater the chance its light—or gravitational waves—will be gravitationally lensed by an intervening galaxy. All together, these circumstances yield a recipe for Smoot and colleagues’ claim LIGO–Virgo must be seeing gravitationally lensed black hole mergers. “We are saying two thirds of their events are lensed,” Smoot says, of LIGO–Virgo’s catalogue of detections.

Daniel Holz, a member of the LIGO collaboration at the University of Chicago is entirely unconvinced. He and his colleagues predicted well before LIGO and Virgo made their detections the observatories would see mergers of black holes of about 30 solar masses each. He agrees greater numbers of low-metallicity stars would have formed in the early universe compared with today’s universe—and hence more 30-solar-mass black holes would have formed then compared with now. But despite most of these bulky black holes forming in earlier cosmic epochs he remains confident LIGO and Virgo are detecting their mergers now, in the relatively local universe, because the gravitational dance that ultimately leads to the coalescence of two orbiting black holes is a process that unfolds over billions of years.

Also, Holz adds, ground-based surveys have shown that some low-metallicity regions do in fact exist in the local universe, all of which could harbor such black hole binaries with 30 solar masses each. “You put all that together and you make a prediction for what you should see with LIGO,” he says. And the detections are in line with the predictions, he adds, making it highly unlikely that any of the LIGO–Virgo events are lensed events. “The current theoretical underpinnings of star formation and evolution, and black hole binary formation and evolution, seem to account for all LIGO observations to date reasonably well. There is no need to go to extremely speculative models.”


Smoot and colleagues, however, are not backing down. They think they have identified at least one definitive lensing event in the LIGO–Virgo data. When a distant source is being lensed, the light or gravitational waves from the source can take multiple paths around the lensing galaxy, and these paths can reach Earth at different times, creating multiple images.

According to their analysis, two events—a detection on August 9, 2017, (GW170809) and another five days later (GW170814)—are actually different images from the same merger. The team argues that the signals share many crucial characteristics, which in both cases lead to nearly identical estimates for the masses of the merging black holes. There is also a small overlap in their approximate locations in the sky.

The LIGO team disagrees. Collaboration member Parameswaran Ajith of the International Center for Theoretical Sciences in Bengaluru, India, and colleagues analyzed all 10 black hole merger events seen by LIGO–Virgo. They looked for consistency between pairs of events, which could be suggestive of lensing. They took into account seven different characteristics of each pair, including a few not considered by Smoot’s team, such as spin angular momentum of the black holes and the orientation of the binary. Two pairs of events, one being the GW170809–GW170814 pair, showed more correlations than others.

But Ajith and colleagues’ analysis showed even for these pairs there was more than a 5 percent chance for the correlations to arise by chance (less than a so-called 2-sigma result). In physics claims of a discovery usually require a 5-sigma result, or less than a 0.00006 percent chance of it being a statistical fluke. “Less than 2-sigma is not even considered plausible,” Ajith says. In other words, the correlations Smoot and his team say are significant may be mere mirages in the data.


Uncontroversial observations of lensed gravitational waves would significantly expand the scope of the science that is possible with LIGO and Virgo. For starters, lensing creates multiple images or signals from the same event, and the arrival of those signals at Earth can be separated by hours or days or weeks. Because the orientation of LIGO’s and Virgo’s instruments relative to the source would have changed between detections, due to Earth’s rotation, this would be akin to having multiple detectors observing (and yielding more information on) the same event. “If you can combine this with optical observations of the lensing galaxy, you might be able to actually pinpoint the binary black hole source very well,” Ajith says.

And because lensed events would be from the early universe, they would allow physicists to ask and answer more nuanced questions about, for example, the evolution of stars and black holes over time. “Instead of doing astronomy, you are also doing cosmology,” Broadhurst says.

Holz is holding out for stronger evidence but is nonetheless excited by the prospect. “Wouldn’t it be awesome, someday, to be able to measure a strongly lensed event?” He thinks it will take the next generation of gravitational wave detectors to unambiguously find ancient black hole mergers, the ripples from which were fortuitously bent toward Earth by unnamed galaxies along the way.