Gravitational waves are notoriously hard to observe — it took 50 years from the first detector prototypes to an actual detection in 2015. But two waves that in January 2020 hit gravitational observatories in Italy and the US have proved more difficult than the average. Scientists spent almost 18 months, longer than for most previous detections, analysing and re-analysing the data, before they were confident enough to announce that signals were indeed what they seemed: ripples in space time caused by black holes swallowing neutron stars.
“It was an event that we had never observed before, and the mass difference between the objects involved was among the highest we had ever seen. We had to do some extra work to create physics models to interpret these events,” says Walter Del Pozzo, a physicist at the University of Pisa and at the VIRGO gravitational observatory in Cascina, near Pisa.
Black holes and neutron stars are both very dense objects that result from the collapse of a star. While the gravity of black holes is so strong that no radiation can escape, neutron stars emit radio waves, X-rays and other electromagnetic signals. No observatory, of any kind, had ever seen them orbit around each other or collide. Now that one such event has been confirmed, scientists expect to see many more when gravitational detectors increase their sensitivity during a series of updates over the next few years.
Three such detectors are active. The Italian/French collaboration EGO manages VIRGO, while the LIGO observatory has twin detectors in the US – one in Livingston, Louisiana, and the other in Hartford, Washington. The three detectors work as one, sharing data and collaborating on their analysis. A fourth detector in Japan, KAGRA, is under development and will add its data in a few years, but its researchers already participate in the scientific work.
So far, these detectors had observed either two black holes or two neutron stars merging, and astrophysicists were eager to also see the third possible combination. “There were few doubts about the existence of binary systems made of a black hole and a neutron star,” says Michela Mapelli, an astrophysicist at the University of Padova and at the National Institute for Astrophysics, who works with VIRGO. “But a part of the community was sceptical that they could actually merge”. For that to happen, says Mapelli, the two objects must orbit very close to each other, resisting the “kick” of the supernova explosion from which the neutron star originates, that pushes them apart.
Joint observation campaigns by LIGO and VIRGO between 2015 and 2019 went by without seeing such a merger. Two events came close, in April and again in August 2019, but the signals were not clear enough. Then on 5 January 2020, during the third observation run, a signal was picked up in Livingston and Cascina (the Hartford detector was off that day), and recognised as a potential black hole/neutron stars mergers by the software protocols that process data in real time. “We have a catalogue of pre-calculated waveforms that tell us what kind of wave to expect from various astrophysical events” says Dal Pozzo. Only ten days later, on 15 January 2020, it happened again, this time with all three detectors up and running.
The scientists then embarked in a painstaking, second-by-second analysis of what was going on at the detector before and after the event, to make sure the waves were not artefacts. “Our signals are drowned in noise” says Giancarlo Cella from the National Institute of Nuclear Physics, who coordinates data analysis at VIRGO. “We can model noise to a certain extent, but there are often exceptional noises that elude simple explanations”. Anything, from seismic activity to a tiny glitch in the laser apparatus, must be factored in. Once the actual signal was isolated, the scientists had to go through the whole list of possible sources of gravitational waves, simulating various versions of them and calculating how likely they were to generate those particular waves.
In a study now published in The Astrophysical Journal Letters1, they ruled out all other explanations and estimated that the 5 January event was caused by the collision of black hole and neutron star, respectively 8.9 times and 1.9 times the mass of our Sun. The objects involved in the 15 January event were slightly smaller, 5.7 solar masses for the black hole and 1.5 for the neutron star. Both events happened between 900 million years and 1 billion years ago.
Scientists had hoped to gain more information by associating an electromagnetic emission, in X-rays or gamma rays for example, to the gravitational waves. But, despite an alert sent out to ground- and space-based observatories immediately after the 15 January event, nothing came up. That could be because the neutron star disappeared into the black hole so quickly that there was no time for its matter to produce a visible burst, or simply because astronomers were looking for a needle in a haystack. “We were looking at such a huge portion of the sky,” notes Mapelli, “that it would have been surprising to find something”.
The scientists also used bespoke statistical models, and a lot of computational power, to estimate how often black holes and neutron stars merge. Relatively often, as it turns out: between 5 and 15 times per year within 1 billion light year from Earth. That does not offer clarity about how black hole/neutron pairs form, which remains an open question. But it means that an increase in sensitivity should soon make detections more frequent. Cella says that VIRGO will double its sensitivity for the next data run, which may begin in the second half of 2022. That will translate into a roughly eight-fold increase in volume of the portion of the Universe explored. Then the sensitivity will double again for the fifth campaign, later in the decade.
More mergers detected will mean a trove of new information for astrophysicists. “For instance, we will be able to understand a mysterious mass gap we observe for compact objects,” says Mapelli. There seem to be very few compact objects between 2 and 5 solar masses, she explains, and it’s not clear whether they can or cannot exist at all. “Exploring this gap will be a blessing for models of formation of black holes and neutron stars,” Mapelli adds. Better sensitivity to high-frequency gravitational waves will also be key, says Cella. “It will allow us to see the final phase of deformation of the neutron star as it falls into the black hole, and that will give us information on the state of the matter inside it,” he explains.A better signal/noise ratio – and one more detector in Japan, when KAGRA starts taking data – will also allow to restrict the sky region where other observatories could find electromagnetic counterparts. Or not find them, which would be just as interesting. “If we continue to not see electromagnetic signals for black hole/neutron star mergers, even with the improved detectors, we will rule them out as sources of short gamma-ray bursts, leaving only neutron star mergers” says Mapelli. Short GRBs are intense electromagnetic signals, lasting less than two seconds, whose origin is still debated.
LIGO and VIRGO will reach their structural limits with their fifth observation campaign, but scientists are already planning for the next jump in sensitivity. In a fitting coincidence, the day after the publication of the new detection the European Commission announced the addition of the Einstein Telescope, a larger and much more sensitive gravitational observatory designed to replace the current ones in the mid-2030s, to a list of future research infrastructures that it plans to fund. Two possible sites are being evaluated, one on the border region between Belgium, Germany, and the Netherlands, and the other near Lula, in north-eastern Sardinia. A decision is expected in 2024.
R. Abbott et al, ApJL, 915 (2021).