2001ApJ...556..340K

We estimate the coalescence rate of close binaries with two neutron stars (NS) and discuss the prospects for the detection of NS-NS inspiral events by ground-based gravitational-wave observatories, such as LIGO. We derive the Galactic coalescence rate using the observed sample of close NS-NS binaries (PSR B1913+16 and PSR B1534+12) and examine in detail each of the sources of uncertainty associated with the estimate. Specifically, we investigate (1) the dynamical evolution of NS-NS binaries in the Galactic potential and the vertical scale height of the population, (2) the pulsar lifetimes, (3) the effects of the faint end of the radio pulsar luminosity function and their dependence on the small number of observed objects, (4) the beaming fraction, and (5) the extrapolation of the Galactic rate to extragalactic distances expected to be reachable by LIGO. We find that the dominant source of uncertainty is the correction factor (up to ≃200) for faint (undetectable) pulsars. All other sources are much less important, each with uncertainty factors smaller than 2. Despite the relatively large uncertainty, the derived coalescence rate is consistent with previously derived upper limits, and is more accurate than rates obtained from population studies. We obtain a most conservative lower limit that the detection rate by LIGO II of about 2 events per year. Our upper limit on the rate is between 300 and 1000 events per year.