GROWTH observations of GW170817: The gravitational wave event GW170817 was caused by the merger of two neutron stars (see the Introduction by Smith). In three papers, teams associated with the GROWTH (Global Relay of Observatories Watching Transients Happen) project present their observations of the event at wavelengths from x-rays to radio waves. Evans et al. used space telescopes to detect GW170817 in the ultraviolet and place limits on its x-ray flux, showing that the merger generated a hot explosion known as a blue kilonova. Hallinan et al. describe radio emissions generated as the explosion slammed into the surrounding gas within the host galaxy. Kasliwal et al. present additional observations in the optical and infrared and formulate a model for the event involving a cocoon of material expanding at close to the speed of light, matching the data at all observed wavelengths. Science , this issue p. 1565 , p. 1579 , p. 1559 ; see also p. 1554

GROWTH observations of GW170817: The gravitational wave event GW170817 was caused by the merger of two neutron stars (see the Introduction by Smith). In three papers, teams associated with the GROWTH (Global Relay of Observatories Watching Transients Happen) project present their observations of the event at wavelengths from x-rays to radio waves. Evanset al.used space telescopes to detect GW170817 in the ultraviolet and place limits on its x-ray flux, showing that the merger generated a hot explosion known as a blue kilonova. Hallinanet al.describe radio emissions generated as the explosion slammed into the surrounding gas within the host galaxy. Kasliwalet al.present additional observations in the optical and infrared and formulate a model for the event involving a cocoon of material expanding at close to the speed of light, matching the data at all observed wavelengths.Science, this issue p.1565, p.1579, p.1559; see also p.1554

I discuss simulations of the coalescence of black hole neutron star binary systems with black hole masses between 14 and 20 \msun. The calculations use a three-dimensional smoothed particle hydrodynamics code, a temperature-dependent, nuclear equation of state and a multi-flavor neutrino scheme. General relativistic effects are mimicked using the \Pacz-Wiita pseudo-potential and gravitational radiation reaction forces. Opposite to previous, purely Newtonian calculations, in none of the explored cases episodic mass transfer occurs. The neutron star is always completely disrupted after most of its mass has been transferred directly into the hole. For black hole masses between 14 and 16 \Msun an accretion disk forms, large parts of it, however, are inside the last stable orbit and therefore falling with large radial velocities into the hole. These disks are (opposite to the neutron star merger case) thin and -apart from a spiral shock- essentially cold. For higher mass black holes ($M_{\rm BH} \ge 18$ \msun) almost the complete neutron star disappears in the hole without forming an accretion disk. In these cases the surviving material is spun up by tidal torques and ejected as a half-ring of neutron-rich matter. None of the investigated systems is a promising GRB central engine. We find between 0.01 and 0.2 \Msun of the neutron star to be dynamically ejected. Like in a type Ia supernova, the radioactive decay of this material will power a light curve with a peak luminosity of a few times $10^{44}$ erg/s. The maximum will be reached about three days after the coalescence and will be mainly visible in the optical/near infrared band. The coalescence itself may produce a precursor pulse with a thermal spectrum of $\sim 10$ ms duration.

We follow the long-term evolution of the dynamic ejecta of neutron star mergers for up to 100 years and over a density range of roughly 40 orders of magnitude. We include the nuclear energy input from the freshly synthesized, radioactively decaying nuclei in our simulations and study its effects on the remnant dynamics. Although the nuclear heating substantially alters the long-term evolution, we find that running nuclear networks over purely hydrodynamic simulations (i.e. without heating) yields actually acceptable nucleosynthesis results. The main dynamic effect of the radioactive heating is to quickly smooth out inhomogeneities in the initial mass distribution, subsequently the evolution proceeds self-similarly and after 100 years the remnant still carries the memory of the initial binary mass ratio. We also explore the nucleosynthetic yields for two mass ejection channels. The dynamic ejecta very robustly produce ‘strong’ r-process elements with A > 130 with a pattern that is essentially independent of the details of the merging system. From a simple model we find that neutrino-driven winds yield ‘weak’ r-process contributions with 50 < A < 130 whose abundance patterns vary substantially between different merger cases. This is because their electron fraction, set by the ratio of neutrino luminosities, varies considerably from case to case. Such winds do not produce any ⁵⁶Ni, but a range of radioactive isotopes that are long-lived enough to produce a second, radioactively powered electromagnetic transient in addition to the ‘macronova’ from the dynamic ejecta. While our wind model is very simple, it nevertheless demonstrates the potential of such neutrino-driven winds for electromagnetic transients and it motivates further, more detailed neutrino-hydrodynamic studies. The properties of the mentioned transients are discussed in more detail in a companion paper.

We study the optical and near-infrared luminosities and detectability of radioactively powered electromagnetic transients (‘macronovae’) occuring in the aftermath of binary neutron star and neutron star black hole mergers. We explore the transients that result from the dynamic ejecta and those from different types of wind outflows. Based on full nuclear network simulations we calculate the resulting light curves in different wavelength bands. We scrutinize the robustness of the results by comparing (a) two different nuclear reaction networks and (b) two macronova models. We explore in particular how sensitive the results are to the production of α -decaying trans-lead nuclei. We compare two frequently used mass models: the finite-range Droplet model (FRDM) and the nuclear mass model of Duflo and Zuker (DZ31). We find that the abundance of α -decaying trans-lead nuclei has a significant impact on the observability of the resulting macronovae. For example, the DZ31 model yields considerably larger abundances resulting in larger heating rates and thermalization efficiencies and therefore predicts substantially brighter macronova transients. We find that the dynamic ejecta from NSNS models can reach peak K-band magnitudes in excess of −15 while those from NSBH cases can reach beyond −16. Similar values can be reached by some of our wind models. Several of our models (both wind and dynamic ejecta) yield properties that are similar to the transient that was observed in the aftermath of the short GRB 130603B. We further explore the expected macronova detection frequencies for current and future instruments such as VISTA, ZTF and LSST.

The detailed observations of GW170817 proved for the first time directly that neutron star mergers are a major production site of heavy elements. The observations could be fit by a number of simulations that qualitatively agree, but can quantitatively differ (e.g., in total r-process mass) by an order of magnitude. We categorize kilonova ejecta into several typical morphologies motivated by numerical simulations, and apply a radiative transfer Monte Carlo code to study how the geometric distribution of the ejecta shapes the emitted radiation. We find major impacts on both spectra and light curves. The peak bolometric luminosity can vary by two orders of magnitude and the timing of its peak by a factor of five. These findings provide the crucial implication that the ejecta masses inferred from observations around the peak brightness are uncertain by at least an order of magnitude. Mixed two-component models with lanthanide-rich ejecta are particularly sensitive to geometric distribution. A subset of mixed models shows very strong viewing angle dependence due to lanthanide “curtaining,” which persists even if the relative mass of lanthanide-rich component is small. The angular dependence is weak in the rest of our models, but different geometric combinations of the two components lead to a highly diverse set of light curves. We identify geometry-dependent P Cygni features in late spectra that directly map out strong lines in the simulated opacity of neodymium, which can help to constrain the ejecta geometry and to directly probe the r-process abundances.