Stephen Smartt and Roy Williams
There are around a dozen binary neutron systems in our galaxy, with the discovery in 1974 of the first example. The Hulse-Taylor binary is a pulsar and a neutron star with an orbital period of around 7.7hrs. The measured decay in the orbital period was spectacularly reproduced with the general relativistic prediction of the system emitting gravitational waves. Russell Hulse and Joseph Taylor were awarded the Nobel prize in 1993 for this discovery. The timing measurements show how the binary period diminishes, a data curve that is exquisitely reproduced by the theory of general relativity and the emission of gravitational waves.
The LIGO and Virgo gravitational wave observatories were built to directly detect gravitational waves from compact binaries. Gravitational waves are emitted by black hole and neutron star binaries with a frequency twice the orbital frequency for, approximately, circular orbits. The Hulse-Taylor binary emits gravitational waves with a frequency around 0.0001 Hz, but this is far below the frequency band that LIGO and Virgo are sensitive too – they are sensitive in the band 10-1000 Hz, so they can only detect binary neutron stars as they approach the last seconds of their merger and their orbital frequency is of this order.
In 2017, the gravitational wave source GW170817 was detected – a highly significant in-spiral and merger signal of a binary neutron star system. This was followed just 1.7 seconds later by a the detection of a weak and short gamma ray burst and the source was localised in the optical to a galaxy at 40Mpc distance. The optical and near infra-red light comes from the thermal emission of a hot expanding ball of gas ejected by the merger. Only around 1% of the mass of the system (about 0.03 solar masses) was ejected, but all of it was likely radioactive heavy isotopes and the decay of these nuclei to stable heavy elements powered this short lived thermal emission. It was a spectacular discovery, but one that has not been repeated. One more high-significance neutron star merger event, GW190425, has been discovered by the LIGO-Virgo-Kagra Science collaboration and one low significance event has been detected. In neither case was there a coincidence GRB, X-ray transient or optical counterpart discovered that is plausibly related to the GW signals. The discovery of GW170817 revealed an astrophysical production source of the heavy elements in the periodic table, proved that gravitational waves travel at the same speed as light, verified the link between neutron star mergers and gamma ray bursts and constrained the equation of state of the matter in neutron stars.
We would like more discoveries of kilonovae! A recent radical change in the field has occurred that may change the way we think of searching for the optical and infra-red emission from merging neutron stars: two long gamma ray bursts in the nearby Universe have been shown to produce optical/infra-red signatures which resemble kilonovae and not supernovae. Up until 2022/23 the idea that long gamma ray bursts all come from collapsing massive stars and supernovae was mostly unchallenged. However GRB211221A and GRB230307A have turned the field on its head. They are both long GRBs, somewhat peculiar in their X-ray properties but sit well within the regime of standard long GRBs in the hardness - duration plot. The kilonova signatures are quite convincing but both were just a little too far to be detected by the LIGO and Virgo detectors. At around 300Mpc they were both relatively close.
All of these events indicate that merging neutron stars exist in the local Universe and can be detected with and without gravitational waves. What will Rubin bring to this rapidly changing field? There are no examples of kilonovae being discovered through optical light alone – even though, since 2017, every wide-field survey on the planet has been looking for them. Many rapidly declining sources, in the vicinity of nearby galaxies have been found but all have been shown to be interlopers, impostors, phonys and fraudsters.
We know they are intrinsically faint, although not much fainter than supernovae. We know they must decline rapidly, since the mass ejected is low and the velocity of ejecta is likely high, around 0.2 to 0.3c. We know now, through the lack of detections since 2017, that they are rare. Which means we must search a large volume for faint and fast decaying sources that have unusual colours. Rubin is well placed although its cadence in its main survey mode (the Wide Fast Deep) is not ideal for objects that have half lives of 1 to 2 days. Rubin can find and detect faint transients – down to ~24 magnitude – meaning it has the potential to detect a kilonova at peak, and roughly 1 mag of decline to a distance of ~1000 Mpc. They will almost certainly be in the data, but identifying them in the optical with no other clues will be challenging. The interlopers will continually pester us. Kilonova are likely (but not always) to turn red quickly and Rubin’s filter set is an advantage. Who will find them and with which broker or combination of brokers?
Lasair has some good tools for finding kilonovae:
-
The Sherlock classification that finds a host galaxy.
-
If there is a host galaxy with a redshift or distance, Lasair computes the extinction-corrected peak absolute magnitude of the object, see here.
-
Fitting an explosion model to the (extinction-corrected) lightcurve, see Bazin black body
-
Utilising the Rubin cadence to find colour. Each object will have two detections in different filters at only 30 minutes apart; an extinction-corrected colour temperature can be derived from this, see Pair feature.
Distance, light-curve evolution and colour evolution are the key aspects. But the challenge is removing the contaminants, to reduce the numbers for spectroscopic and multi-wavelength followup. It is a purity versus completeness challenge. Machine learning may help, but by the time we have sufficient data on an object to give a reliable score, it may be gone and the opportunity lost.
It’s fifty years since discovery of binary neutron stars and ten years since the discovery of gravitational waves. We hope for momentous discoveries in 2026.