“The signal caught our eye immediately”
The discovery of gravitational waves on 14 September 2015 was the crowning moment of a search which had lasted decades and employed ingenious methods. The Max Planck Institute for Gravitational Physics, with its branches in Potsdam-Golm and Hanover, played a crucial role in this success. This is where researchers are working not only on innovative technologies but also on theoretical models, virtual simulations and data analysis. We discussed this work and the importance and consequences of the discovery with Directors Bruce Allen, Alessandra Buonanno and Karsten Danzmann.
Mr Allen, Ms Buonanno, Mr Danzmann: As members of an international network of gravitational wave detectors, the LIGO Virgo collaboration, you and the staff at your Institute played a significant role in the very first measurement of gravitational waves. Congratulations!
All three: Thanks very much!
Did you expect to make the discovery at this point in time?
Karsten Danzmann: No, not at all. It was a complete surprise. In the middle of September 2015, the American LIGO detectors – designed along the lines of a Michelson interferometer like our GEO600 detector – were still only in test mode after undergoing a rather long upgrade phase. The plan was for the scientific measuring operation to begin a few days later. The scientists were still checking that the instruments were working as planned. And they were. But that they were operating so well and would be able to receive a gravitational wave signal right at the start – nobody had expected that.
Bruce Allen: The signal arrived late morning on 14 September 2015 Central European Time. In the US it was nighttime and our colleagues there were asleep. So it was two members of the Max Planck Institute for Gravitational Physics who were the first to look at their screen a few minutes after the detectors had been triggered. They analyzed the data for several hours and then sent an initial e-mail to the collaboration. We couldn’t believe it at first, especially since the signal was so strong and looked so perfect that we first asked ourselves whether it actually was real.
Karsten Danzmann: It must be said here: we regularly simulate the impact of gravitational waves on the detectors for test purposes. This allows us to test the operability of the instruments and also to check the detection chain and establish that the scientists are working independently of each other.
Bruce Allen: In the first few weeks after the discovery we did indeed have doubts that someone could have mistakenly injected an artificial signal or forgotten to tell us about it. We invested a lot of effort into excluding this possibility. But in the end we knew: the signal originates from outer space. We had become witnesses to the fact that, in a distant galaxy, two black holes had collapsed into one another!
What does such a signal look like?
Alessandra Buonanno: The signal swept through the LIGO detectors for half a second. It looked remarkably simple! A sine wave of about 10 to15 cycles whose amplitude first grew, then reached a maximum and eventually died out. Meanwhile, its frequency increased more and more, ending up at a constant value. This characteristic signal can be explained as follows: As the two black holes orbit each other, they lose energy because of gravitational wave emission. Therefore, they come closer and closer until they collide with each other and merge, forming a more massive black hole. It then rings like a bell for a little while before settling down. Before merger, the signal’s frequency is proportional to the orbital frequency, and its amplitude to the characteristic velocity of the binary, which is almost the speed of light during the last stages of the evolution. Once the new black hole forms, it rings down emitting gravitational waves at a constant frequency.
Bruce Allen: I didn’t expect that the very first detection would enable us to deduce the event so directly from the waveform. I had assumed that the first detections would be much weaker and that we would need our analytical programs to fish them out of the data. And also that it would be difficult to understand what is really happening there. The fact that it stands out so clearly in the raw data, and is even visible to the human eye, is remarkable.
Even though the gravitational wave signal in this case is obviously easy to see with the naked eye: it is imperative to undertake a sound data analysis. How is this done and what is the role of the Max Planck Institute for Gravitational Physics?
Karsten Danzmann: While the detectors are running, the measurement data are automatically trawled continuously for signals. If something is found, the scientists are sent an e-mail.
Bruce Allen: The foundations for the algorithm which tracked down the latest signal were laid by colleagues at the University of Florida. This algorithm searches the LIGO detectors for a deflection at the same frequencies so that the events in both detectors match. In our Working Group we have spent years extending and improving this code in order to filter out from the data the special signals from binary systems with black holes of moderate mass. These improvements were one of the reasons that the latest event was discovered. And as far as the algorithm for the precision analysis applied after the detection is concerned, the colleagues at our Institute belong to one of only two groups of experts in the world.
Where are the calculations done?
Bruce Allen: Most of them are done on the ATLAS computer cluster here in Hanover. It has roughly the same capacity as the rest of the collaboration together has available.
Karsten Danzmann: After excluding all other external perturbations including earthquakes, for example, the signal is compared with synthetically generated waveforms. We thus determine the properties of the astrophysical source emitting the gravitational waves.
How are those wave signals modelled and implemented into your search?
Alessandra Buonanno: First, we developed sophisticated, analytical approximations to describe the two-body dynamics and gravitational radiation during the phase in which two black holes come ever closer to each other. Then, we used the results of numerical-relativity simulations of binary black holes to model the merger and ring down signal. It is not possible to employ only numerical-relativity waveforms in searches and follow-up analyses because it takes one month or more to simulate the last 15 orbits of a binary black-hole coalescence. The waveform models we have developed were also implemented and employed in the continuing search for binary coalescences in LIGO's data. This search observed the black-hole merger known as GW150914 with high enough significance to be confident in the detection.
And afterwards, you can say precisely what the system you have found really looks like?
Alessandra Buonanno: Having identified the signal in the data, we used our waveform models to run follow-up analyses and infer the astrophysical properties of the source. We found that the binary system was composed of two black holes which had 36 and 29 solar masses. The two black holes merged into a single, rotating black hole with a mass of around 62-times the mass of the Sun. The binary was 1.3 billion light years away. Furthermore, the signal being quite loud, it was also possible to employ our waveform models to look for violations of Einstein’s theory of general relativity. No deviations were found!
Apart from the signal strength: was this system a surprise in any other way?
Alessandra Buonanno: We did not know if black holes with masses larger than 20 solar masses existed, but we did know that if they existed, they would be the strongest gravitational wave sources for LIGO. They were the golden sources we always dreamed! This is because, for such massive binary black holes, the merger signal lies in the most sensitive region or sweet spot of the detector, and it is at merger that the signal is strongest.
Which frequencies are these?
Karsten Danzmann: Between 60 and 250 hertz. In this range, the LIGO detectors are meanwhile almost ten times as sensitive as before the upgrade. This is something which we are particularly delighted about, incidentally: almost all the developments that have made Advanced LIGO so much more sensitive were developed or tested out at GEO600. At higher frequencies where we expect the signals from two fusing neutron stars, for example, the instruments are currently a factor of three better than before. In the coming months this is still to be increased to a factor of ten, however. At very low frequencies, the dominance of the seismic effects is too great. But, in future, this gap will be closed by the VIRGO detector in Italy, which is also a member of our network. Its technology is also currently being modernized and is to resume operation next year.
And what is the situation with GEO600 in Ruthe near Hanover?
Karsten Danzmann: It is smaller, and so at low frequencies it is not sensitive enough for such signals. Its strength lies at higher frequencies. But the main thing we have here is decades of tradition in developing technologies. All innovations which have had their origins here can meanwhile be found in the other detectors in the network; they include special systems for suspending mirrors, and also the laser technology and the optical layout of the interferometer in general. We provided the hardware for the pre-stabilized laser systems used with Advanced LIGO. Advanced LIGO is our detector as well!
The discovery has shown that the calculation that the new measurement sensitivity of the detectors would finally enable us to measure gravitational waves directly, proved to be correct. And even earlier than was hoped for. Which further developments and observations do you expect in the near future?
Bruce Allen: In the immediate future it could become particularly exciting. We have now observed one system very well. I estimate that during the next six months of scientific operation following a further, brief update phase over the course of the year we will see a system like this every three or four days. Towards the end of the next measuring period we will have around 20 such detections. We will be able to see what the mass spectrum of such systems is. And we will learn something about the evolution of such systems, because some of them will be close, others more distant, which means they formed at an earlier point in time. This will tell us something about the proportion of heavy elements in the universe during the various eras, for example, because this has a considerable impact on the formation rate of massive stars and black holes in particular. And then we naturally hope to also find all the other types of sources which are to be expected – the fusing of two neutron stars or combinations of a neutron star and a black hole.
What does the discovery mean for physics in the broader sense?
Karsten Danzmann: I think it has enormous significance for physics and astronomy. Not so much because gravitational waves have finally been detected. Nobody doubted they would be! But because gravitational wave astronomy has now become mainstream astronomy. We suddenly have a new tool at our disposal with which to study the dark side of the universe. We have to realize that more than 99 percent of space emits no light and no electromagnetic radiation. All we know about this part at the moment is that it is subject to gravitation. The great hope now is that it will be possible to investigate it.
Bruce Allen: But first and foremost we have shown that we can measure gravitational waves directly. We can now use this to do research. And we are now in a position to test the General Theory of Relativity in strong gravitational fields. Until now, it has mainly proved that Einstein’s theory is completely correct. So I don’t think this will tell us something fundamentally new about physics that we don’t already know. But we have a wonderful method for checking these laws.
Alessandra Buonanno: It is such an enormous discovery that it is difficult to immediately anticipate all the repercussions for gravity, fundamental physics and astrophysics, but its echoes will be reverberating in those fields for many, many years. It’s fantastic that the announcement takes place shortly after the 100-year anniversary of Einstein’s publication of the paper where he predicted the existence of gravitational waves! We now have a new tool to probe the universe and unveil its dark, most extreme sides. We have discovered that stellar-mass black holes exist, that they exist in pairs – i.e., in binary systems – and that they can be quite massive. And yes, the observation of binary black-hole mergers provides us with the remarkable opportunity to see how gravity operates at such extreme conditions and test whether Einstein’s theory of general relativity still holds. So far, so good!
Bruce Allen: I also think back to the centenary anniversary of the General Theory of Relativity which we celebrated in autumn 2015 in Berlin: because even Einstein himself did not believe that it would ever be possible to measure gravitational waves since they are so weak. Neither did he believe in black holes. We have shown that he was wrong on both counts. But I don’t think this would have bothered him. I think he would have been delighted!
Interview: Felicitas Mokler