Article Index

Gravitational-wave sources and multi-messenger time domain Astrophysics

The launch of THESEUS will coincide with a golden era of multi-messenger astronomy. With the first detection of gravitational waves (GWs) by Advanced detectors (Abbott et al. 2016a, Abbott et al. 2016b), a new window on the Universe has been disclosed. By the end of the present decade, the GW sky will be routinely monitored by a network of second generation GW detectors, an ensemble of Michelson-type interferometers composed by the two Advanced LIGO (ALIGO) detectors in the USA and by Advanced Virgo (AdVirgo) in Italy, plus ILIGO in India and KAGRA in Japan joining later within a few years. By 2030, more sensitive third generation GW detectors, such as the Einstein Telescope and Cosmic Explorer, are planned to be operative. In parallel to this advancement, IceCube and KM3nNeT and the advent of Megaton detectors will likely revolutionize neutrino astrophysics.

  • 8

  • 8b

  • 8c

  • 8d


Figure 6. Top left: by the end of the ‘20s, if ET will be a single detector, almost no directional information will be available for GW sources, and a GRB-localising satellite will be essential to discover EM counterparts. If multiple third-generation detectors will be available, the typical localization uncertainties will be comparable to the aLIGO ones. The plot shows the SXI field of view (~110x30 deg2) superimposed on the probability skymap of GW151226 observed by aLIGO. Top right: THESEUS will detect and localize down to 0.5-1 arcmin the soft X-ray short/long GRB afterglows, of NS-NS (BH) mergers and of many classes of galactic and extra-galactic transients. For several of these sources, THESEUS/IRT will provide detection and study of associated NIR emission, location within 1 arcsec and redshift. Middle: the detection, study and arcsecond localization of afterglow and kilonova emission  from short GRB/GW events will be possible with THESEUS/IRT. In this figure, we show in particular the light curve of the kilonova associated to the gravitational wave/short GRB event GW170817/GRB170817A in the IRT filters. The continuous and dashed red lines indicate the THESEUS/IRT limiting H magnitudes for imaging and prism spectroscopy, respectively, with 300s of exposure. Bottom: THESEUS will be able to promptly and accurately localize e.m. counterparts to GW events.


Several of the most powerful transient sources of GWs predicted by general relativity, e.g. binary neutron star (NS-NS) or NS-black hole (BH) mergers (with likely detection rate of ~50 per year by the LIGO-Virgo network at design sensitivity; Abadie et al. 2010), are expected to produce bright electromagnetic (EM) signals across the entire EM spectrum and in particular in the X-ray and gamma-ray energy bands, as well as neutrinos. GW detectors have relatively poor sky localization capabilities of ~10 sq. degrees (e.g. Klimenko et al. 2011). Neutrino detectors such as IceCube and KM3NeT can localize to an accuracy of better than a few sq. degrees but within a smaller volume of the Universe (e.g. Santander 2016 and references therein). In order to maximize the science return of the multi-messenger investigation it is essential to have an in-orbit trigger and search facility that can either detect an EM signal simultaneous with a GW/neutrino event or rapidly observe with good sensitivity the large error boxes provided by the GW and neutrino facilities following a trigger.

This combined requirements are uniquely fulfilled by THESEUS, which is able to trigger using XGIS or SXI and observe a very large fraction of the GW/neutrino error boxes within an orbit due to the large grasp of the SXI instrument (if compared for example to current generation X-ray facilities such as Swift, THESEUS/SXI has a grasp of ~40 times that of Swift/XRT). For events triggered on-board with XGIS or SXI, GW searches can also be carried on the resultant known sky locations with lower GW detector signal-to-noise thresholds and hence an increased search distance.

The detection of EM counterparts of GW (or possibly neutrino) signals will enable a multitude of science programmes (e.g. Bloom et al. 2009; Phinney 2009) by allowing for parameter constraints that the GW/neutrino observations alone cannot fully provide. For example, finding a GW/neutrino source EM counterpart in X-rays with THESEUS/SXI, will allow to localize the source with an accuracy good enough for optical follow-up and hence to possibly measure its redshift and luminosity. On the other hand, not finding an EM counterpart will constrain merger types (such as BH-BH mergers), emission mechanisms, astrophysical conditions at the time of merger, and total energetics.

Moreover, in 2020s synergies with the future facilities like James Webb Space Telescope (JWST), WFIRST, ATHENA, Einstein Probe, E-ELT, TMT, GMT, SKA, CTA zPTF and LSST telescopes would be desirable in order to be able to extend the multi-messenger observations across the whole EM spectrum.

Gravitational wave transient sources

NS-NS / NS-BH mergers: Collimated EM emission from Short GRBs

Compact binary coalescing (CBC) mergers are among the most promising sources of GWs that will be detected in the next decade. Indeed, these systems are expected to radiate GWs within the most sensitive frequency range of ground based GW detectors (1-2000 Hz), with large GW energy output, of the order of 10-2 Moc2, and gravitational waveforms well predicted by General Relativity (see Baiotti & Rezzolla 2016 for a review). The expected rate of NS-NS systems inferred from binary pulsar observations and population synthesis modeling, is taken to lie between 10 and 10000 Gpc-3 yr-1. To date, no NS-BH systems have been observed, but the rate can still be predicted through population synthesis, constrained by the observations of NS-NS, to be ~1-1000 Gpc-3 yr-1 (see Abadie et al. 2010 and references therein).

Mounting indirect evidence associates short GRB progenitors to CBC systems with at least one neutron star (e.g. Berger 2014) and provides possible hints that the merger of magnetized NS binaries will lead to the formation of a jet with an opening angle of ~30°-40° (Rezzolla et al. 2011). GW observations combined with multi-wavelength follow-up campaigns in the next years will likely confirm this scenario by simultaneous detections of CBC-GWs temporally and spatially consistent with short GRBs. With the present sensitivity distance range, of the second-generation GW detectors can detect CBC systems up to ~200-300 Mpc. Within such a volume the expected short GRB rate is rather low, of the order of ~1/yr or less. However, by the time of the launch of THESEUS, gravitational radiation from such systems will be easily detectable by third-generation detectors such as the Einstein Telescope (ET) up to redshifts z ~ 2 (e.g. Chassande-Mottin et al. 2011) implying that a large fraction of the short GRBs that THESEUS will detect, i.e., ~20 short GRBs per year, will have a detectable GW emission. The estimated rate of detected short GRBs is ~1-10 Gpc-3 yr-1 (Wanderman & Piran 2015 and reference therein, see also Ghirlanda et al. 2016).

Almost all short GRBs are accompanied by an X-ray afterglow that SXI will detect and monitor just after the burst emission. Once localized with SXI, about 40% of detected short GRBs are expected to have a detectable optical/IR counterpart. The IRT could point the SXI localized afterglow within few minutes from the trigger. If bright enough, spectroscopic observations could be performed onboard, thus providing redshift estimates and information on chemical composition of circumburst medium. In addition, precise sky coordinates will be disseminated to ground based telescopes to perform spectroscopic observations.

Distance measurements of a large sample of short GRBs, combined with the absolute source luminosity distance provided by the CBC-GW signals, can deliver precise measurements of the Hubble constant (Schutz 1986), helping to break the degeneracies in determining other cosmological parameters via CMB, SNIa and BAO surveys (for example Dalal et al. 2006). Using the predicted XGIS/SGI GRB rates, the XGIS field of view and assuming a conservative 20% beaming fraction, the predicted THESEUS/ALIGO/AdVirgo triggered coincidence number is 3-4 within the nominal mission life, but rises to 10 or more when including SXI follow-up observations of ALIGO/AdVirgo error boxes. With 10 GW+EM events, the Hubble constant could be constrained to 2-3%, thus providing a precise independent measure of this fundamental parameter (Dalal et al. 2006). In addition, each individual joint GW+EM observation would provide an enormous science return from THESEUS. For example, the determination of the GW polarization ratio would constrain the binary orbit inclination and hence, when combined with an EM signal, the jet geometry and source energetics. Likewise, a better understanding of the NS equation of state can follow from combined GW and EM signals (e.g., Lasky et al. 2014, Ciolfi & Siegel 2015a,b, Messenger et al. 2015).

NS-NS / NS-BH mergers: Optical/NIR and soft X-ray isotropic emissions

Nearly isotropic EM emission is expected from NS-NS / NS-BH mergers at minutes-days time scale from the merger. GW emission depends only weakly on the inclination angle of the inspiral orbit, and GW detectors will mostly trigger on off-axis mergers (i.e. for binary system with a non zero inclination angle). Re-pointing of THESEUS in response to a GW trigger will allow SXI to study off-axis X-ray emission. One expected off-axis X-ray emission is the late afterglows from the laterally spreading jet as soon as it decelerates (“orphan afterglows”, van Eerten et al. 2010). Peak brightness is expected at 1-10 days after the trigger, with peak fluxes equal or below ~10-12-10-13 erg cm-2 s-1 at ~200 Mpc (Kanner et al. 2012). Therefore, despite their low collimation, off-axis afterglows will be detected only for the most nearby CBC systems.

Another nearly-isotropic emitting component is expected if a massive millisecond magnetar is formed from two coalescing NSs. In this case, X-ray signals can be powered by the magnetar spindown emission reprocessed by the matter surrounding the merger site (isotropically ejected during and after merger), with luminosities in the range 1043-1048 ergs/s and time scales of minutes to days (Metzger & Piro 2014, Siegel and Ciolfi 2016a,b). Alternatively, X-ray emission may come from direct dissipation of magnetar winds (e.g. Zhang 2013, Rezzolla & Kumar 2015). Numerical simulations suggest that such emission is collimated but with large half-opening angles (30°-40°, beaming factor of ~0.2). As an additional channel in X-rays, the magnetar may power the isotropically expelled matter through wind pressure to relativistic speeds generating a shock with ISM (“confined winds”). Synchrotron radiation produced in the shock is emitted nearly isotropically, with an enhanced intensity near the equator. A beaming factor of ~0.8 is expected in this case (e.g. Gao et al. 2013). Overall, typical time scales for these X-ray signals are comparable to magnetar spin down time scales of ~103-105 s, and the predicted luminosities span a wide range that goes from~1041 erg/s to ~1048 erg/s. With THESEUS/SXI in combination with sensitive GW detectors as ET, several X-ray counterparts of GW events from NS-NS merging systems will be easily detected possibly up to large distances (depending on the largely uncertain intrinsic luminosity of such X-ray component), providing a unique contribution to probe and classify X-ray emission from BNS systems.

In the optical band, the expected EM component is represented by the so-called “macronova” (often named “kilonova”) emission. During NS-NS or NS-BH mergers, in fact, a certain amount of ejected mass is expected to become unbound. This matter has the unique conditions of high neutron density and temperature to initiate the r-process nucleosynthesis of very heavy elements. Days after merger, the radioactive decay of such elements heats up the ejected material producing a transient signal peaking in the optical/near-infrared (NIR) band and with luminosities of ~1041 erg/s (e.g. Li & Paczynski 1998, see also Baiotti & Rezzolla 2016 for a review). Today, macronovae are promising electromagnetic counterparts of binary mergers because (i) the emission is isotropic and therefore the number of observable mergers is not limited by beaming; (ii) the week-long emission period allows for sufficient time needed by follow-up observations. Once identified, source location can be accurately recovered, allowing for the identification of the host galaxy and the search for counterparts in other electromagnetic bands (e.g. radio). The detectability of macronovae is currently limited by the lack of sufficiently sensitive survey instruments in the optical/NIR band that can provide coverage over tens of square degrees, the typical area within which GW events will be localized by the Advanced LIGO-Virgo network (LIGO Scientific Collaboration et al. 2013). In the 2020s we expect several synergies at different wavelegths such as: 1) the space-based telescopes James Webb Space Telescope (JWST), ATHENA and WFIRST, 2) the ground-based telescope with large FOV like zPTF and LSST which will be able to select the GW candidates in order to follow-up them afterwards 3) the large multi-wavelengths telescopes such as the 30-m class telescopes GMT, TMT and E-ELT, which will all follow-up the optical/NIR counterparts like macronovae, and 4) the Square Kilometer Array (SKA) in the radio, which is well suited to detect, for example, the late-time (~weeks) signals produced by the interaction of the ejected matter with the interstellar medium. THESEUS/IRT will be perfectly integrated in this search, thanks to its photometric and spectroscopic capabilities and its spectral range coverage, allowing us to improve the study of such (until now) uncertain emission mechanisms. With IRT we will acquire not only the light curves of EM counterparts but also their spectra, thus having the opportunity to disentangle the different components associated to those events (e.g. disk wind + dynamical ejecta contributions).

Core collapse of massive stars: Long GRBs, Low Luminosity GRBs and Supernovae

The collapse of massive stars are expected to emit GWs since a certain degree of asymmetry in the explosion is inevitably present. Estimates of the GW amplitudes are still largely uncertain and the expected output in energy has enormous uncertainties, ranging from 10-8 to 10-2 Moc2. If the efficiency of the GW emission is effectively very low, then only the third-generation GW detectors such as ET, which will be operative at the same time as THESEUS, will be able to reveal the GW emission from these sources and thus obtain crucial insights on the innermost mass distribution, unaccessible via EM observations.

The collapsar scenario invoked for long GRBs (e.g. Woosley 1993, Paczynski 1998) requires a rapidly rotating stellar core, so that the disk is centrifugally supported and able to supply the jet. This rapid rotation may lead to non-axisymmetric instabilities, such as the fragmentation of the collapsing core or the development of clumps in the accretion disk. Inspiral-like GW signals are predicted with an amplitude that will be observable by ET to luminosity distances of order 1 Gpc (e.g. Davies et al. 2002). With a rate of observed long GRB of ~0.5 Gpc-3 yr-1 , THESEUS could provide a simultaneous EM monitoring of ~1 long GRBs per year. Off-axis X-ray afterglow detections (“orphan afterglows”) can potentially increase the simultaneous GW+EM detection rate by a factor that strongly depends on the jet opening angle and the observer viewing angle. The possible large number of low luminosity GRBs (LLGRBs) in the nearby Universe will provide clear signatures in the GW detectors because of their much smaller distances with respect to long GRBs.

GW/EM synergy is key in this investigation. Since the gamma-ray emission and afterglow are expected to be produced at large distances (i.e. ~1013 cm) from the central engine, they provide only indirect evidence for the nature of that engine. By contrast, gravitational waves will be produced in the immediate vicinity of the central engine, offering a direct probe of its physics.

Soft Gamma Repeaters

Fractures of the solid-crust on the surface of highly magnetized neutrons stars and dramatic magnetic-field readjustments represent the most widely accepted explanation to interpret X-ray sources such as giant flares and soft gamma repeaters (SGRs) (e.g. Thompson & Duncan 1995). NS crust fractures have also been suggested to excite non radial oscillation modes that may produce detectable GWs (e.g., Corsi & Owen 2011, Ciolfi et al. 2011). The most recent estimates for the energy reservoir available in a giant flare are between 1045 erg (about the same as the total EM emission) and 1047 erg. The efficiency of conversion of this energy to GWs has been estimated in a number of recent numerical simulations and has been found to be likely too small to be within the sensitivity range of present GW detectors (Ciolfi & Rezzolla 2012, Lasky et al. 2012). However, at the typical frequencies of f-mode oscillations in NSs frequencies, ET will be sensitive to GW emissions as low as 1042 – 1044 erg at 0.8 kpc, or about 0.01% to 1% of the energy content in the EM emission in a giant flare. In the region of 20 − 100 Hz, ET will be able to probe emissions as low as 1039 erg, i.e. as little as 10−7 of the total energy budget (e.g. Chassande-Mottin et al. 2010 and reference therein).

Neutrino sources

Several high-energy sources that THESEUS will monitor are also thought to be strong neutrino emitters, in particular SNe and GRBs. The shocks formed in the GRB ultra-relativistic jets are expected to accelerate protons to ultrarelativistic energies and that, after interacting with high energy photons, produce charged pions decaying as high energy neutrinos (>105 GeV, e.g. Waxman & Bachall 1997). Pulses of low energy neutrinos (<10 MeV) are expected to be released during core-collapse supernovae (CCSNe) with an energy release up to 1053 erg. Low energy neutrinos have been detected from SN1987A at 50 kpc distance and are expected in general from CCSNe. Still significant uncertainties are affecting supernova models. GW and neutrino emission provide important information from the innermost regions as the degree of asymmetry in the matter distribution, as well the rotation rate and the strength of the magnetic fields, that can be used as priors in numerical simulations (e.g. Chassande-Mottin et al. 2010 and reference therein).

Because of the very small cross-sections and low fluxes of the emitted neutrinos, neutrino detectors necessarily require huge amounts of water or liquid scintillator in neutrino detectors. Future Megatons detectors as Deep-TITAND that are expected to work during the 3rd generation GW detectors, will reach distances up to 8 Mpc, that would guarantee simultaneous GW/neutrino and EM detection of ~ 1 SN per year. Very promising for such multi-messenger studies are the LLGRBs, given their expected larger rate than for standard long GRBs (up to 1000 times more numerous) and their proximity. For long GRBs, in the context of CCSNe, joint THESEUS and GW/neutrino observations would constrain progenitor models, clarifying the fraction of energy channeled via dynamical instabilities (Fryer et al., 2002) and the relative neutrino/EM energy budgets. Neutrino observations would also constrain the composition of the GRB jet and the relation of GRBs to high-energy cosmic rays (Abbasi et al. 2012).