Targeted search for the stochastic gravitational-wave background from the galactic millisecond pulsar population

TL;DR

This study conducts a targeted search for the gravitational-wave background from galactic millisecond pulsars using LIGO and Virgo data, setting upper limits on pulsar properties and projecting future detector sensitivities.

Contribution

It presents the first likelihood-based targeted search for the SGWB from galactic millisecond pulsars, incorporating population synthesis models and providing new upper limits on pulsar ellipticity.

Findings

No significant SGWB signal detected.

Upper limits on the number of observable pulsars and their average ellipticity.

Projected sensitivities for future GW detectors.

Abstract

The millisecond pulsars, old-recycled objects spinning with high frequency $\mathcal{O}$ (kHz) sustaining the deformation from their spherical shape, may emit gravitational waves (GW). These are one of the potential candidates contributing to the anisotropic stochastic gravitational-wave background (SGWB) observable in the ground-based GW detectors. Here, we present the results from a likelihood-based targeted search for the SGWB due to millisecond pulsars in the Milky Way, by analyzing the data from the first three observing runs of Advanced LIGO and Advanced Virgo detector. We assume that the shape of SGWB power spectra and the sky distribution is known \textit{a priori} from the population synthesis model. The information of the ensemble source properties, i.e., the in-band number of pulsars, $N_{\textrm{obs}}$ and the averaged ellipticity, $\mu_\epsilon$ is encoded in the maximum likelihood statistic. We do not find significant evidence for the SGWB signal from the considered source population. The best Bayesian upper limit with $95\%$ confidence for the parameters are $N_{\textrm{obs}}\leq8.8\times10^{4}$ and $\mu_\epsilon\leq1.4\times10^{-6}$, which is comparable to the bounds on mean ellipticity with the GW observations of the individual pulsars. Finally, we show that for the plausible case of $N_{\textrm{obs}}=40,000$, with the one year of observations, the one-sigma sensitivity on $\mu_\epsilon$ might reach $1.5\times10^{-7}$ and $4.1\times10^{-8}$ for the second-generation detector network having A+ sensitivity and third-generation detector network, respectively.