**3. CBC Multi-Messenger Astrophysics Today**

It is often assumed that the multi-messenger revolution has a very precise starting date: 17 August 2017, with the detection of GW170817 by the LIGO/Virgo interferometers and short GRB170817 by the Fermi/GBM and INTEGRAL/SPI. Unfortunately, no other event was detected simultaneously by GWs and gamma-rays (or any electromagnetic radiation) during the third LIGO/VIRGO observing run (O3). Since one swallow does not make a summer, all questions presented in the previous section are still open. Certainly, GW/GRB1708017 has shown how powerful the multi-messenger approach could be. Its full impact, however, measured by the ability to open a long-lasting, brand-new field must still be determined, and it will greatly depend on the capability of collecting statistical samples of GW-electromagnetic events in the near future, that is, during the LIGO/VIRGO O4 and O5 observing runs planned for this decade.

As of now, the LIGO/Virgo interferometers have detected two NS-NS coalescence events (GW170817 and GW190425) and one very likely BH-NS merger (GW190814).

**The GW/GRB170817** event does not need to be commented on in detail here; many comprehensive reviews exist [6]. Briefly, Fermi/GBM and INTERGAL/SPI detected a short burst lasting for about 2 s, just 1.7 s after the NS-NS detection by LIGO/Virgo. The LIGO/Virgo error box was about 30 deg<sup>2</sup> , a value that did not improve much after including the Inter Planetary Network (IPN) error box. The distance of the source 40+<sup>8</sup> <sup>−</sup>14Mpc was determined directly from the GW detection and greatly limited the search for an optical counterpart to only about 50 galaxies. An optical counterpart was discovered in NGC4993 only after about 11 h from the GW/GRB event, which gave rise to an impressive follow-up by nearly one hundred ground-based and space-based telescopes, placing constraints on the optical/IR counterpart of the GW source/GRB and discovering the first confirmed kilonova in history. The GRB detection, which was nearly simultaneous with the GW signal, confirmed that, at least in this case, a merger of two NSs is the origin of short GRBs. The follow-up observations allowed the complex jet structure to be determined, and the jet was observed as off-axis for the first time [10–12]. It was also found that kilonovae were one of the prime sites of heavy element production through the r-processes [6,13].

**GW190425** was detected with a high significance by one interferometer only, and thus the uncertainty region included a large fraction of the sky [14]. No GRB was detected at the same time, with an upper limit on the 50–300 keV fluence of <sup>∼</sup> <sup>10</sup>−<sup>6</sup> ergs/cm<sup>2</sup> in 1 s [15], corresponding to a few ph/cm<sup>2</sup> , by INTEGRAL/SPI. Swift/BAT and Fermi/GBM covered only a fraction of the region, and therefore could not contribute much to the INTEGRAL observation. The distance of the event inferred from the GW signal is 159+<sup>69</sup> <sup>−</sup><sup>71</sup> Mpc. Therefore, even for a gamma-ray luminosity comparable to that of GRB170817, its flux would need to be 16 times dimmer, <0.1 ph/cm<sup>2</sup> ; hence, it is relatively difficult to detect with the available detectors. Moreover, the observed gamma-ray flux is a strong function of the inclination at which the jet is observed (a power law with an exponent −5 ÷ −6). Unfortunately, the inclination of this binary system was not well constrained by the interferometers, but assuming similar characteristics of GRB170817, we can infer a lower limit on the inclination between the jet and the line-of-sight of about 10 deg.

**GW190814** was detected by all three interferometers, thus providing good constraints on the position of the source (error box of 18.5 deg<sup>2</sup> ), distance (241+<sup>41</sup> <sup>−</sup><sup>45</sup> Mpc) and inclination (46 +/− 11 deg), in addition to the masses of the two coalescing objects and of the remnant (this was the GW event with the most unequal mass ratio, 0.11 +\− 0.01 ever detected). No significant electromagnetic counterpart was reported for this event, with INTEGRAL/SPI inferring a three-sigma upper limit for the flux of <sup>3</sup> <sup>×</sup> <sup>10</sup>−<sup>7</sup> erg/cm2/s in the energy range 75–2000 keV [16]. Swift/BAT covered >99% of the error box with >10% partial coding, reporting a five-sigma upper limit of 10−<sup>7</sup> erg/cm<sup>2</sup> between 15–350 keV in 1 s [17]. The lack of gamma-ray detection is likely due to the high inclination of the system, which corresponds to a huge (factor of 106–10<sup>7</sup> ) reduction in the observed flux. The jet emission at these inclinations is hardly observable by any conceivable all-sky monitor at the moment.

In summary, the lessons learnt from the three NS-NS and BH-NS events collected so far are: (1) the need for an **all-sky monitor** with a sufficient sensitivity to detect off-axis jets with intrinsic luminosities down to 10<sup>47</sup> ergs, at least to a distance of 100–200 Mpc and an inclination of 10–20 deg (of course smaller inclination values allow for detection over larger distances). In fact, within this decade the LIGO/Virgo interferometers will reach their target sensitivity and the searched volume will become much larger with respect to O1–O3: the horizon for NS-NS merging events detected with a signal-to-noise ratio of 8 will reach ~200 Mpc for LIGO and 100–130 Mpc for Virgo in O4, implying a discovery volume ~100 times larger than in the GW170817 case. The capability to instantaneously cover the whole sky is mandatory, because the number of events will be small; therefore, missing simply one event would mean a considerable loss for scientific research. (2) The capability of determining the position of the transients with uncertainties is smaller than a few degrees. Within the volume defined by this spatial constraint and the distance of the GWE provided by the interferometric measurements, the number of optical transients will be small enough to easily assess the correct transient to associate with the GWE, thus prompting further follow-ups.
