Artificial intelligence is allowing scientists to see the sources of gravitational wave faster and more accurately than ever before. Credit: James Josephide
Following the recent overwhelming success of deep learning and artificial intelligence in several fields of research, industry and medicine, researchers from the ARC Centre of Excellence of Gravitational Wave Discovery (OzGrav) and the University of Western Australia (UWA), including PhD student and the paper’s first author Chayan Chatterjee, have built a deep learning model using an Artificial Neural Network to pinpoint where in the sky gravitational wave signals have come from. The model can localise the source of gravitational waves produced by colliding pairs of black holes potentially as much as a thousand times faster than any other technique.
Professor Amitava Datta, a scientist from UWA who contributed to the study, says: ‘This work is a very interesting example of learning patterns from simulated data for predicting the outcome of real events, in this case the location of gravitational wave sources. Perhaps this approach using deep learning will be more and more useful in basic sciences in general.’
Data intensive astronomy expert Kevin Vinsen from the International Centre for Radio Astronomy Research says: ‘This project is an excellent example of how a multi-disciplinary approach can solve the problem’.
The basic structure of an Artificial Neural Network. The circles represent the neurons or nodes and the arrows represent connections between one neuron to another. Credit: Chayan Chatterjee
Gravitational waves are small ripples in the space-time continuum caused by colossal stellar events such as colliding black holes. In September 2015, following recent advances in detector sensitivity, the LIGO Scientific Collaboration detected gravitational waves for the first time. This was a landmark achievement in human discovery and heralded the birth of the new field of gravitational wave astronomy.
The need for speed and accuracy is particularly important in the context of gravitational wave localisation—scientists need to tell a global network of telescopes where to point on the sky as quickly as possible, so they can see any electromagnetic light that may also have come from the gravitational wave event. The current algorithm used to locate gravitational wave sources in real time takes a few seconds to process. More accurate methods usually take several hours to compute. The light generated by gravitational wave events can be very short-lived at certain wavelengths, like short gamma ray bursts, which last a mere 2-3 seconds, so scientists need methods that can rapidly process huge data as fast and accurately as possible.
The idea behind deep learning is simple: it’s an algorithm designed to mimic the functioning of neurons in our brain to carry out tasks, like categorising observed stimuli. This is done by making the network learn the correlations between a labelled input dataset and the output it is trying to predict. Just like electric signals or synapses flow through neurons in our brain, the input information provided to an Artificial Neural Network travel through layers of nodes (usually several layers deep), with each layer introducing some non-linearity to the input. This non-linearity helps the network learn complex features of the data. The ‘learning’ happens through a rigorous ‘training’ of the network. During the training, the predictions of the network are compared with the true values, and the parameters of the network are adjusted to minimise any erroneous gaps.
In their recently published paper, Chatterjee and the team from UWA successfully trained the Artificial Neural Network to learn the input data for source localisation. The data was pre-processed to extract the important physical Physics parameters from simulated gravitational wave signals, injected into ‘random noise’. The network classified these signals into several classes and accurately identified the source direction of the gravitational waves. The model localised the test samples much faster than other methods and at a low computational cost. The researchers plan to extend this work for pairs of merging neutron stars and neutron star-black hole systems.
Chatterjee says: ‘Hopefully the methods we introduce can also be translated to other areas of research and industry and help further untap the seemingly limitless potential of deep learning and artificial intelligence’.
OzGrav’s Chief Investigator Professor Linqing Wen who led the study says: ‘The future is wide open for gravitational wave discovery using the machine learning technique’.
This paper is an outcome of a multi-disciplinary collaboration of UWA’s Gravitational Wave Astronomy Group led by OzGrav’s Chief Investigator Professor Linqing Wen, data intensive astronomy expert Kevin Vinsen from International Centre for Radio Astronomy Research (ICRAR), and Professor Amitava Datta of UWA’s Computer Science and Software Engineering.
Link to publication: https://journals.aps.org/prd/abstract/10.1103/PhysRevD.100.10302
Artist's impression of the binary neutron star merger producing GW190425. Credit: National Science Foundation/LIGO/Sonoma State University/A. Simonnet.
A new collaborative study with the ARC Centre of Excellence for Gravitational Wave Discovery (OzGrav) reveals a possible collision of two neutron stars earlier in 2019—only the second time this type of cosmic event had ever been detected. The gravitational-wave observatory network, that includes the National Science Foundation's LIGO and the European Virgo detectors, picked up what appeared to be gravitational ripples from a collision of two neutron stars back on 25 April 2019.
Gravitational waves and light were first witnessed in the same event in 2017. This second event in 2019, called GW190425, did not result in any light being detected; however, researchers have learned that the collision resulted in a merged object with an unusually high mass.
OzGrav postdoctoral researcher Simon Stevenson says: ‘This event is a perfect example of how gravitational-wave astronomy is a completely new and unique way of looking at the Universe. Binaries with similar masses to this event may not exist in the Milky Way or may be completely invisible to conventional radio telescopes’.
Neutron stars are the remnants of dead stars that exploded. When two neutron stars spiral together, they undergo a violent merger that sends gravitational waves shuddering through the fabric of space and time.
The gravitational waves first detected in 2015 were generated by the fierce collision of two black holes. Since then, scientists have registered dozens of new candidate black hole mergers. The first detection of a neutron star merger took place two years later, in 2017.
OzGrav Postdoctoral Researcher Vaishali Adya says: ‘This detection manifests the importance of continued improvement of the already amazingly sensitive gravitational wave detectors, as this event would not have been observable prior to the latest upgrades. OzGrav played a vital role in these upgrades, one of which involved reducing the quantum noise in the detectors’.
OzGrav Postdoctoral Researcher Xingjiang Zhu says: ‘The combined mass of the merging objects is surprisingly high, much greater than any previously known double neutron star binaries including the one detected in 2017. This provokes us to think about the nature of this event and how the source might have been formed’.
The combined mass of the merged bodies in this event is about 3.4 times that of the mass of the Sun. Typically, neutron star collisions are only known to happen between pairs of neutrons stars with a total mass up to 2.9 times that of the Sun. One possibility for the unusually high mass is that the collision took place not between two neutron stars, but a neutron star and a black hole, since black holes are heavier than neutron stars. But if this were the case, the black hole would have to be exceptionally small for its class. Instead, the scientists believe it is more likely that the event was a shattering of two neutron stars and that their merger resulted in a newly formed black hole.
Neutron star pairs are thought to form either early in life—when companion massive stars successively die one by one—or when they come together later in life within dense, busy environments. The data from the 2019 event do not indicate which of these scenarios is more likely—more data and new models are needed to explain the unexpectedly high mass. The discovery suggests that we may have detected an entirely new population of binary neutron star systems.
OzGrav Associate Investigator Greg Ashton says: ‘This event was really interesting. The chirp-like signal was seen by two of the three detectors for about 128 seconds before the final merger. Unfortunately, one of the detectors was not observing at the time, which meant that the sky localization was poor. Perhaps because of this, and because it was so far away, no electromagnetic light was observed from this event. Nevertheless, we saw it very clearly in the gravitational wave data and could use that to calculate the masses, spins, and orientations of the objects’.
‘Additional exciting and unexpected discoveries can be expected as the sensitivity of the LIGO detectors improves. OzGrav is working closely with LIGO to improve their sensitivity, developing new instrumentation and analysis techniques’, says Professor Peter Veitch, University of Adelaide OzGrav Node Leader
The results were announced today at the American Astronomical Society meeting in Honolulu, Hawaii.
The full scientific article will be available here post-embargo: https://dcc.ligo.org/P190425/public