Seconds instead of hours: AI locates merging neutron stars faster
Merging neutron stars give themselves away through gravitational waves. However, there is not much time left to observe the kilonova. An AI could soon help.
Artistic representation of the fusion of two neutron stars
(Image: MPI for Intelligent Systems / A. Posada)
A research team led by the Max Planck Institute for Intelligent Systems has developed an AI-based method that can use gravitational waves to determine much more quickly and precisely where in the sky signals from a merger of two neutron stars are likely to be found. While the characterization of such a collision based on gravitational waves has so far taken an hour in the fastest case, it only takes around a second with the new technology. This acceleration could be crucial for quickly aligning telescopes and observatories to the respective region of the sky in future and being able to directly observe such a process more frequently. So far, this has required luck.
The "holy grail"
As the Max Planck Society now reminds us, only black holes have a higher mass density than neutron stars. But while a merger of two black holes only emits gravitational waves, such a process in neutron stars is also accompanied by an extreme flash of light across almost the entire electromagnetic spectrum. The aim is to locate the source before the flash of light can be detected on Earth. However, depending on how far away from us the merger takes place, there is not much time to do this. Current gravitational wave detectors therefore usually only provide a warning time of a few minutes. Thanks to the new AI method, this could be enough.
Videos by heise
"A fast and accurate analysis of gravitational wave data is crucial in order to localize the source and align telescopes as quickly as possible and observe all associated accompanying signals," explains study leader Maximilian Dax. According to the team, the new method could one day help to observe electromagnetic signals produced before and during the merger of two neutron stars. This could provide new insights into the merging process and provide answers to previously unsolved questions. At the same time, the expensive observation time could be used more efficiently.
The algorithm, named Dingo-BNS ("Deep Inference for Gravitational-wave Observations from Binary Neutron Stars"), could thus become a central component of so-called multi-messenger astronomy. This is the term used to describe the exploration of the universe with detectors for completely different signals, some of which have only been operational for a few years. The kilonovae, which are created when neutron stars merge, are considered the "holy grail of multi-messenger astronomy", writes the Max Planck Society – . Many elements are created in the process. The algorithm for real-time localization is presented in the scientific journal Nature.
(mho)
