Article révisé par les pairs
Résumé : The cosmic evolution of the neutron-star merger (NSM) rate can be deduced from the observed cosmic star formation rate. This allows us to estimate the rate expected in the horizon of the gravitational wave detectors advanced Virgo and advanced LIGO and to compare those rates with independent predictions. In this context, the rapid neutron-capture process, or r-process, can be used as a constraint assuming NSM is the main astrophysical site for this nucleosynthetic process.We compute the early cosmic evolution of a typical r-process element, europium. Eu yields from NSM are taken from recent nucleosynthesis calculations. The same approach allows us to compute the cosmic rate of Core-Collapse SuperNovae (CCSNe) and the associated evolution of Eu. We find that the bulk of Eu observations at [Fe/H] > -2.5 can be rather well fitted by either CCSN or NSM scenarios. However, at lower metallicity, the early Eu cosmic evolution favours NSM as the main astrophysical site for the r-process. A comparison between our calculations and spectroscopic observations at very low metallicities allows us to constrain the coalescence time-scale in the NSM scenario to ~0.1-0.2 Gyr. These values are in agreement with the coalescence time-scales of some observed binary pulsars. Finally, the cosmic evolution of Eu is used to put constraints on (i) the NSM rate, (ii) the merger rate in the horizon of the gravitational wave detectors advanced Virgo/ad LIGO, as well as (iii) the expected rate of electromagnetic counterparts to mergers ('kilonovae') in large near-infrared surveys.