Search for continuous gravitational waves from ten H.E.S.S. sources using a hidden Markov model

TL;DR

This paper reports a search for continuous gravitational waves from ten pulsars associated with H.E.S.S. sources using a hidden Markov model, analyzing aLIGO data to identify potential signals and distinguish them from noise.

Contribution

It introduces a semicoherent search method combining the $ ext{F}$-statistic with a hidden Markov model to improve detection of wandering quasi-monochromatic signals from young pulsars.

Findings

Identified 12 candidates surviving data quality vetoes.

Analyzed 1-Hz sub-bands around key frequencies of each pulsar.

Detected candidates require further observation to confirm or refute.

Abstract

Isolated neutron stars are prime targets for continuous-wave (CW) searches by ground-based gravitational$-$wave interferometers. Results are presented from a CW search targeting ten pulsars. The search uses a semicoherent algorithm, which combines the maximum-likelihood $\mathcal{F}$-statistic with a hidden Markov model (HMM) to efficiently detect and track quasi$-$monochromatic signals which wander randomly in frequency. The targets, which are associated with TeV sources detected by the High Energy Stereoscopic System (H.E.S.S.), are chosen to test for gravitational radiation from young, energetic pulsars with strong $\mathrm{\gamma}$-ray emission, and take maximum advantage of the frequency tracking capabilities of HMM compared to other CW search algorithms. The search uses data from the second observing run of the Advanced Laser Interferometer Gravitational-Wave Observatory (aLIGO). It scans 1$-$Hz sub-bands around $f_*$, 4$f_*$/3, and 2$f_*$, where $f_*$ denotes the star's rotation frequency, in order to accommodate a physically plausible frequency mismatch between the electromagnetic and gravitational-wave emission. The 24 sub-bands searched in this study return 5,256 candidates above the Gaussian threshold with a false alarm probability of 1$\%$ per sub-band per target. Only 12 candidates survive the three data quality vetoes which are applied to separate non$-$Gaussian artifacts from true astrophysical signals. CW searches using the data from subsequent observing runs will clarify the status of the remaining candidates.