Characterization of the stochastic signal originating from compact binary populations as measured by LISA

Nikolaos Karnesis, Stanislav Babak, Mauro Pieroni, Neil Cornish, and Tyson Littenberg

Phys. Rev. D 104, 043019 – Published 20 August 2021

Abstract

The Laser Interferometer Space Antenna (LISA) mission, scheduled for launch in the early 2030s, is a gravitational wave observatory in space designed to detect sources emitting in the millihertz band. In contrast to the present ground-based detectors, the LISA data are expected to be a signal dominated, with strong and weak gravitational wave signals overlapping in time and in frequency. Astrophysical population models predict a sufficient number of signals in the LISA band to blend together and form an irresolvable foreground noise. In this work, we present a generic method for characterizing the foreground signals originating from a given astrophysical population of coalescing compact binaries. Assuming idealized detector conditions and a perfect data analysis technique capable of identifying and removing the bright sources, we apply an iterative procedure which allows us to predict the different levels of foreground noise.

Received 22 April 2021
Accepted 29 July 2021

DOI:https://doi.org/10.1103/PhysRevD.104.043019

Physics Subject Headings (PhySH)

Gravitational wave detection Gravitational wave sources Gravitational waves

Gravitation, Cosmology & Astrophysics

Authors & Affiliations

Nikolaos Karnesis^1,2, Stanislav Babak^1,3, Mauro Pieroni ⁴, Neil Cornish⁵, and Tyson Littenberg⁶

¹APC, AstroParticule et Cosmologie, CNRS, Université de Paris, F-75013 Paris, France
²Department of Physics, Aristotle University of Thessaloniki, Thessaloniki 54124, Greece
³Moscow Institute of Physics and Technology, Dolgoprudny, Moscow region, Russia
⁴Blackett Laboratory, Imperial College London, SW7 2AZ, United Kingdom
⁵eXtreme Gravity Institute, Department of Physics, Montana State University, Bozeman, Montana 59717, USA
⁶NASA Marshall Space Flight Center, Huntsville, Alabama 35812, USA

Article Text (Subscription Required)

Click to Expand

References (Subscription Required)

Click to Expand

Issue

Vol. 104, Iss. 4 — 15 August 2021

Reuse & Permissions

Access Options

Article Available via CHORUS

Download Accepted Manuscript

Author publication services for translation and copyediting assistance advertisement

Images

Figure 1
(a) Illustration of the progress of the procedure for estimating the confusion noise. For the population of the CGBs described in Sec. 3, it takes 6–10 iterations for the algorithm to converge, depending on the convergence tolerance criterion. The starting data are represented with the gray curve (300 segment averaged spectrum), while the final combined instrument and confusion noise is represented here in bright magenta. The dark blue dashed line represents the analytic model fit of the confusion noise component, together with the associated error bars. In this example, the observation time is $T_{obs} = 2 yr$ . (b) Using the analytical model of Eq. (7) and the parameter values from Table 2, we can make a prediction of the level of the confusion noise due to CGBs, depending on a given observation time. Here, we depict the power spectrum of the residual data with gray, while the colored dashed lines represent the model prediction for the given $T_{obs}$ .
Reuse & Permissions
Figure 2
The 2D projections of the posterior distributions of the parameters of model of Eq. (6), as sampled with MCMC methods (see the text for details). The fit was performed on a dataset with duration of 10 yr, after subtracting the bright sources with SNR threshold $ρ_{0} = 5$ .
Reuse & Permissions
Figure 3
Fit of the two models we considered to the confusion residuals signal from CGBs. The first model, following faithfully Eq. (6), assumes a fixed spectral index of $- 7 / 3$ in strain units or $n_{s} = 2 / 3$ in $Ω_{GW}$ units (shown in blue). The pink data refer to a model where the spectral index $n_{s}$ is a free parameter of the fit (see the text for details).
Reuse & Permissions
Figure 4
Left: the residual confusion noise created by a population of SBBHs for the SNR thresholds $ρ_{0} = {5, 15}$ . Both curves have been estimated using a running median on the data. Right: the number of resolvable sources that are detected and subtracted from the data for different values of SNR threshold $ρ_{0}$ .
Reuse & Permissions
Figure 5
Left: the total signal of the unresolved sources from both populations of CGBs and SBBHs (shown in blue), after subtracting the brightest sources. The signal plus instrumental noise is shown with gray color. Right: the number of subtracted sources for different assumptions about the SNR threshold for the SBBHs, $ρ_{0, sbbh}$ (represented with the pink bars). For the given population and observation duration, it is fairly evident that for $ρ_{0, sbbh} > 15$ there are not SBBHs recovered from the data. The number of CGB sources that were subtracted (blue) is the same for all cases, indicating that the total SBBH residual confusion signal is not strong enough to affect their recovery. Here, we assumed $ρ_{0, cgb} = 7$ .
Reuse & Permissions
Figure 6
Histograms of the recovered errors for the parameters of the subtracted CGBs. The optimal case, where there are only CGBs in the data, is depicted with blue, while the relative errors of the same population with the presence of SBBHs is shown in pink. The same type of histograms are plotted for the brighter ( $ρ > 20$ ) sources in green and light blue, respectively.
Reuse & Permissions

Physical Review D

covering particles, fields, gravitation, and cosmology