Advanced VIRGO: detector optimization for gravitational waves by inspiralling binaries

An advanced European detector is currently considered. A strong R&D program will be necessary to improve the actual sensitivity of VIRGO and GEO. Naturally such effort must aim to optimize detection, given limited budget and manpower. Among sources, coalescing binaries appear among the most interesting and known. Thus, improvements in signal to noise ratio (S/N) by noise abatement in VIRGO, for inspiralling binaries of small chirp mass, are discussed. VIRGO detection bandwidth is divided into four sub-bands according to the dominant noise (whether seismic, thermal from pendula and mirrors, shot). Firstly, it is identified that mirror thermal noise reduction provides the highest gain increment in S/N. On the assumption that the technological challenges lying ahead, and the relative necessary budgets, are equally relevant in all sub-bands, it is recommended to gear R&D strategies to a future EURO detector operating at cryogenic temperatures. Secondly, the frequency of 50 Hz is identified as the single optimal frequency at which lies the potential largest increment in S/N for VIRGO. Thirdly, enlargement of the bandwidth, where noise is reduced, produces a shift to higher frequency of the optimal frequency (e.g. for 100 Hz band, the optimal frequency is 75 Hz). Fourthly, noise reduction factor and width of band, where such reduction occurs, both concur to S/N improvement and equivalent gains may be achieved, proposing alternative R&D strategies. Finally, on the basis of our recent estimates on coalescing Neutron Stars (NS) binaries rates, the detection rate by the actual VIRGO configuration is about one every 27 years. An improvement in the strain sensitivity of a factor 3 in the mirror thermal noise band, implies a detection rate of about one coalescing NS binary every 5 years, or 4 detection a year for a noise reduction factor of 10.