The latest results from LIGO, Virgo, and KAGRA bring several key revelations, including that black holes tend to come with certain masses.

Earlier this week, the vast international team associated with three gravitational-wave projects unveiled the results from their latest observing run. The collaboration had already published two key detections from that run, the first-ever smashups of black holes with neutron stars. But this long-awaited third catalog adds significantly to researchers’ full tally, raising the total number of gravitational-wave events to 90.

The catalog includes compact objects caught colliding during the second half of the collaboration’s third observing run (called O3b), which ran from November 2019 until March 2020. This run included both Europe’s Virgo and the U.S.-based LIGO detectors; the Japanese KAGRA project joined the fun for the campaign’s last two weeks.

All four detectors use lasers bounced off mirrors to measure infinitesimal changes in distance as spacetime scrunches and stretches when a gravitational ripple passes through. The observatories turn up thousands of potential events, which scientists weed through with complex computer algorithms.

a range of lines with blue dots on top and bottom creating a shape similar to a bird with spread wings. red and yellow dots run along the outline of the bottom half, all on a black background.
Scientists have now detected 90 gravitational-wave events, created by the merger of either black holes, neutron stars, or both. The dots indicate the masses of the objects that merged and of the object they created. Pink and yellow dots are detections from electromagnetic observations.
Credit: LIGO-Virgo / Aaron Geller / Northwestern University

Starting with O3, the LVK Collaboration issues public alerts for gravitational-wave events, to enable rapid response by astronomers looking for fleeting glows from the neutron stars or black holes that merged. (Pairs of black holes shouldn’t flash, unless they’re surrounded by gas.) O3b issued 39 alerts, not one of which came with light.

The new analysis combs through the data more carefully than those initial passes. It kicks out about half of the 39 and adds 17 more events that escaped earlier detection, for a total of 35. Combined with the recently revised list of previous detections (which changed the tally from 50 to 55), that gives us a total of 90.

As with previous runs, the new catalog contains mostly black hole mergers — 32 of the total 35 pairs were binary black holes. But there are also the two neutron star–black hole collisions, and one event of indeterminant type: It might have been a black hole gnashing a neutron star, but chances are the smaller object was just a tiny black hole. (And by tiny, I mean about 2.8 times the Sun’s mass.)

Not all of these events are assured to be real; the team estimates that about 10% are false alarms, given their generous inclusion of all events with more than a 50% chance of being legit. Even so, scientists now have nearly 100 examples of crashing objects creating waves in the fabric of spacetime — that’s spectacular given that six years ago, we’d detected zero.

Revelations about Neutron Stars and Black Holes

The new catalog, called GWTC-3, contains several notable events. One crash involved one of the lightest neutron stars ever detected by any method, at 1.2 solar masses. Another involved eye-catchingly hefty black holes (about 87 and 61 Suns) that add to astronomers’ discomfort (more on that later).

This video shows cartoon animations of each of the binaries caught merging during the O3b observing run.
Zoheyr Doctor / CIERA / LIGO-Virgo Collaboration

The list also includes black holes that were spinning like tops in the same direction as their circuit around each other, an alignment more likely to happen if the stars that died to make the black holes were paired since birth. But there was also at least one binary in which the black holes were spinning upside-down compared with their orbit.

Yet as regular S&T readers may know, what interests me most is the big picture. In addition to the detailed catalog, the LVK Collaboration released three other papers, one of which is a 60-page population-level analysis. This kind of study is about statistics, not individual events. The researchers used the catalog’s 76 most reliable events and looked at what they tell us as a cohort.

The population study reveals several interesting discoveries, three of which caught my eye:

First, there’s a clear drop in the number of objects just above 2 solar masses. Astronomers had predicted they wouldn’t see objects between roughly 3 and 5 solar masses, due to physical limitations on how big a neutron star can be as well as observations of binary systems in our galaxy. But the gravitational-wave data don’t show a hard upper edge to this putative gap, nor does it appear to be totally empty. Neutron stars can’t be above about 2.3 Suns, the data indicate, so maybe black holes can be smaller than we thought they could. Previous results support that idea.

Second, not all black holes are created equal. If you look at the larger of the two black holes involved in each merger, these primaries cluster around three different masses: 10, 17-ish, and 35 Suns. Those with 10 solar masses have a good explanation: Black holes with masses this low are unlikely to pair up after formation in a star cluster, so the binaries might all have been remnants of stars that were born and died as fraternal twins. But it’s unclear why some black holes would more often have masses of 17 or 35 Suns.

Third, there’s no evidence yet for an upper mass gap. This result is definitely something to pay attention to. Astronomers have long predicted that there should be a dearth of black holes between roughly 50 and 120 solar masses, because stars large enough to make black holes that big will tear themselves apart when they die, leaving no remnant. Oversized black holes from the previous observing run had already made astronomers squirm. But although the latest gravitational-wave data do show a drop-off at masses above 45 Suns or so, it’s not precipitous. Nor have we seen any black holes above 120 Suns, so there’s no upper edge to the predicted gap. Calculations suggest that if there is a gap, it starts above 75 solar masses — far higher than expected.

Perhaps the unexpectedly large black holes don’t come from normal stars. Instead, they might be second-generation black holes made from mergers, or they could have beefed up thanks to gas they scarfed down. These different scenarios would cause the black holes to spin in certain ways, but so far the spin measurements don’t obviously support any one theory.

LVK will return for the fourth observing run in late 2022, when further upgrades may increase the number of detections by a factor of three. We might be seeing alerts five times a week!


The LIGO Scientific Collaboration, the Virgo Collaboration, and the KAGRA Scientific Collaboration. “GWTC-3: Compact Binary Coalescences Observed by LIGO and Virgo During the Second Part of the Third Observing Run.”

The LIGO Scientific Collaboration, the Virgo Collaboration, and the KAGRA Scientific Collaboration. “The population of merging compact binaries inferred using gravitational waves through GWTC-3.”

You can also find reader-friendly summaries of these research papers on LIGO’s outreach page, in multiple languages.


Image of Bob


November 12, 2021 at 9:33 pm

Thanks for the update!
Questions: does the horizontal axis in "Masses in the Stellar Graveyard" have meaning? If so, what? There's certainly a beautiful symmetry in the displayed data.

Thanks, Bob W

You must be logged in to post a comment.

Image of ljnester


November 15, 2021 at 9:25 am

Can a region of with a group of black holes collapse without a merger?
This idea from a short story got me thinking.
I found the following formula for the event horizon radius: R= (2GM)/(cc).

If I am understanding this right, if you had 4 black holes of the same mass placed at the vertexes of a triangular pyramid with its edge 3 times the radius of the event horizon of one of the black holes, then the combined mass of the cluster would have an event horizon of radius of 4 times one of the holes.
The whole cluster would then be in this event horizon without a merger.
So, for me the idea is possible.

Could some larger black holes get bigger this why?

You must be logged in to post a comment.

You must be logged in to post a comment.