Astron. Astrophys. 336, 786-790 (1998)

1. Introduction

Gravitational waves (GW) from unresolved binary stars in the Galaxy forms a confusion limit in the frequency domain of LISA laser space interferometer (Bender et al. 1996) Hz that effectively adds to the detector's noise. Different aspects of this background have been studied earlier by Rosi and Zimmerman (1976), Lipunov & Postnov (1987), Lipunov, Postnov & Prokhorov (1987), Hils, Bender & Webbink (1990), Bender & Hils (1997), Giampieri & Polnarev (1997), Postnov & Prokhorov (1998).

The importance of the knowledge of binary confusion limit within the LISA frequency band is dictated by many factors - for example, it limits the angular resolution of the detector (Cutler 1997) and restricts the possibilities of observing possible relic GW backgrounds (see, e.g., discussion in Grishchuk 1997). Since LISA with its three arms is expected to operate as essentially single interferometer (Schutz 1997), it can only see noise sources above its intrinsic noise . In our previous paper (Postnov & Prokhorov 1998) we focused on the stochastic GW noise from galactic binaries and showed that, depending on the (unknown) galactic binary white dwarf merger rate, this noise becomes lower than the expected LISA rms noise at frequencies Hz. ¹ We concluded that to be detectable by LISA, relic GW backgrounds should be as high as at Hz.

Clearly, the galactic binary GW noise, tracing the distribution of stars in the Galaxy, should be highly anisotropic (of order of one magnitude higher in the direction of the Galactic center; see calculations by Lipunov et al. 1995) and thus modulated with LISA turning in its orbit. This distincitve feature of the galactic stochastic background can in fact be used to measure it by space-born interferometers (Giampieri & Polnarev 1997). This is not the case for a GW-background produced by extragalactic binaries, which becomes important at frequencies where galactic binaries do not form continuous noise ( Hz for 1-year integration).

A crude estimate shows that the isotropic extragalactic background is expected to be an order of magnitude smaller than the average GW noise from galactic binaries (see Lipunov et al. (1987, 1995), Hils et al. (1990)). All estimates of this background made so far, however, have not taken into account the fact of a strong global star formation evolution recently revealed by different astronomical observations (see the results of Canada-France-Hawaii survey of far galaxies with (Lilly et al. 1996), the analysis of galaxies with Lyman jump at (Madau et al. 1996), and the analysis of the HST deep field survey of galaxies with (Connoly et al. 1997)). According to these studies, the global star formation rate strongly increases with redshift, peacking at (see Fig. 1 below reproduced from Connoly et al. (1997), with points from Madau et al. (1996) corrected for extinction as in Boyle & Terlevich (1997)).

In a Euclidean space, the remote sources would contribute more and more to the energy density with distance r, as , leading to the photometric paradox; in a FRW expanding Universe this is not the case. However, the effects of proper density evolution of galaxies (star formation) with redshift, together with the proper evolution of the sources inside galaxies, could compensate for the cosmological energy density dilation. So a priori it is not at all clear if the remote extragalactic binaries can be neglected in studying extragalactic binary GW background. This is the motivation to our calculations.

© European Southern Observatory (ESO) 1998

Online publication: July 20, 1998
helpdesk.link@springer.de