All-sky search for continuous gravitational waves from isolated neutron stars in the early O3 LIGO data

We report on an all-sky search for continuous gravitational waves in the frequency band 20–2000 Hz and with a frequency time derivative in the range of $[-1.0,+0.1]\times {10}^{-8}\mathrm{Hz}/\mathrm{s}$. Such a signal could be produced by a nearby, spinning and slightly nonaxisymmetric isolated neutron star in our Galaxy. This search uses the LIGO data from the first six months of Advanced LIGO’s and Advanced Virgo’s third observational run, O3. No periodic gravitational wave signals are observed, and 95% confidence-level (C.L.) frequentist upper limits are placed on their strengths. The lowest upper limits on worst-case (linearly polarized) strain amplitude $h_0$ are $\sim 1.7\times {10}^{-25}$ near 200 Hz. For a circularly polarized source (most favorable orientation), the lowest upper limits are $\sim 6.3\times {10}^{-26}$. These strict frequentist upper limits refer to all sky locations and the entire range of frequency derivative values. For a population-averaged ensemble of sky locations and stellar orientations, the lowest 95% C.L. upper limits on the strain amplitude are $\sim 1.4\times {10}^{-25}$. These upper limits improve upon our previously published all-sky results, with the greatest improvement (factor of $\sim 2$) seen at higher frequencies, in part because quantum squeezing has dramatically improved the detector noise level relative to the second observational run, O2. These limits are the most constraining to date over most of the parameter space searched.

Figure 1

Comparison in frequency and spin-down for this search with those of previous all-sky searches of LIGO O2 data. The shaded rectangle with vertical bars shows the 20–2000 Hz and $-{10}^{-8}–{10}^{-9}\mathrm{Hz}/\mathrm{s}$ range for this O3a search. The slightly larger rectangle with horizontal bars shows the region searched in the O2 data with the Frequency Hough method [12, 13]. The smaller rectangle with crossed diagonal bars shows the region searched by the distributed-computing project Einstein@Home [14]. The solid line at zero spin-down depicts the specialized O2 search for low-ellipticity millisecond pulsars using the Falcon method [15, 16, 17] (the thickness of the line overstates the coverage in spin-down range). The dotted curves indicate contours of constant equatorial ellipticity $\epsilon =({10}^{-8},{10}^{-7},{10}^{-6},{10}^{-5}\mathrm{and}{10}^{-4})$ for a star with stellar spin-down dominated by gravitational wave emission.

Figure 2

Illustration of powerflux upper limit validation for very low, low, mid and high frequencies. Each point in the upper left, upper right, lower left and lower right panels represents a separate injection in the 20–60 Hz, 60–475 Hz, 475–1475 Hz and 1475–2000 Hz frequency ranges, respectively. Each established upper limit (vertical axis) is compared against the injected strain value defined by the horizontal axis (diagonal line defines equality of upper limit obtained and true injected strain). These injections are for random polarizations. Corresponding scatter plots for linearly polarized injections confirm the desired 95% coverage, regardless of polarization.

Figure 3

Injection (software simulations) recovery efficiencies in the 20–60 Hz, 60–475 Hz, 475–1475 Hz and 1475–2000 Hz frequency bands are shown in the upper left, upper right, lower left and lower right panels, respectively, for stages 0, 1 and 2 of the search. The injected strain divided by the 95% C.L. upper limit in its band (without injection) is shown on the horizontal axis. The percentage of surviving injections is shown on the vertical axis, with a horizontal dashed line drawn at the 95% level. The vertical dashed line marks a relative strain of unity. Ideally, the recovery efficiencies should lie to the left of the vertical line or above the horizonal line. One observes, however, that the ideal recovery efficiency is significantly degraded below 60 Hz, where large instrumental artifacts in the H1 data and only modest expected Doppler modulations make clean signal recovery more challenging. Most of the degradation stems from the band below 27 Hz for which H1 line contamination is severe and for which the H1 noise floor is substantially higher than the L1 noise floor.

Figure 4

Upper limits on gravitational strain amplitude for this O3a analysis and for previous O2 analyses. The dark blue curve shows worst-case (linearly polarized) 95% CL upper limits in analyzed (62.5 mHz, 125 mHz, 250 mHz) sub-bands for the broad (20–475 Hz, 475–1475 Hz, 1475–2000 Hz) search bands. The lowest, purple curve shows upper limits assuming a circularly polarized source. The middle (green) curve shows approximate population-averaged all-sky upper limits inferred from the circularly polarized limits. For comparison, the uppermost yellow-green diamonds show previous population-averaged 95% CL all-sky upper limits derived from LIGO O2 data using the Frequency Hough method [12, 13]. The orange curve shows 90% CL upper limits from the O2 Einstein@Home search [14] over the band 20–585.15 Hz and over a spin-down range that is 26% of the range used here. The light blue curve shows 95% CL upper limits from the O2 Falcon search [15, 16, 17] over the 20–2000 Hz band and a severely restricted spin-down range. See Fig. 1 for a comparison of parameter space coverage from these various searchers.

Figure 5

Ranges (kpc) of the powerflux search for neutron stars spinning down solely due to gravitational radiation (“gravitars”) under different assumptions. The three sets of three curves (purple, green dotted, yellow-green diamonds) that generally rise with frequency are the ranges for which the O3a circular-polarization, O3a population-averaged and O2 population-averaged Frequency Hough [12, 13] upper limits apply for three different assumed equatorial ellipticities of $\epsilon ={10}^{-4}$, ${10}^{-6}$ and ${10}^{-8}$. In each case, the curve stops at the frequency at which the implied spindown magnitude at the corresponding range exceeds the maximum range allowed by the maximum spindown magnitude (${10}^{-8}\mathrm{Hz}/\mathrm{s}$) used in these analyses. The three black curves (solid, dashed, dot-dashed) that peak below 100 Hz represent the maximum possible ranges for which the O3a circular-polarization limits apply under the assumption that the maximum spindown magnitude is $({10}^{-8},{10}^{-10},{10}^{-12})\mathrm{Hz}/\mathrm{s}$. Searching for high spindown magnitudes simultaneously probes higher ellipticities and larger regions of the galaxy, reaching well beyond the galactic center ($\sim 8.5\mathrm{kpc}$) at high ellipticities and low frequencies. More realistically, a neutron star’s spin-down may be dominated by electromagnetic radiation, in which case searching for high spindown magnitude achieves a shorter search range, but enables detection of stars that would not be detected under the gravitar assumption. (color online).

Figure 6

Shaded regions (yellow) overlaid on the run-averaged spectra (inverse-noise weighted) from the H1 (magenta) and L1 (blue) O3a datasets indicate vetoed bands for which outlier follow-up is impeded by artifacts. The widest affected bands correspond to test-mass “violin modes” (ambient vibrations of the silica fibers from which LIGO mirrors are suspended) near multiples of 500 Hz. Top panel: Full 20–2000 Hz band. Bottom panel: Magnification of example 1900–2000 Hz band, for which line artifacts are dominated by the 4th harmonics of violin modes. (color online).

Figure 7

Example of a “strain histogram” graph used in vetoing outliers surviving stage-2 follow-up for which instrumental contamination is apparent. The solid curves show the O3a-run-averaged H1 (red) and L1 (blue) amplitude spectral densities in a narrow band containing an artifact at 1740.39 Hz. The dotted curves show histograms of expected strain excess from H1 (black) and L1 (magenta) signal templates added to smooth backgrounds interpolated from neighboring frequency bands. In this depiction, the strain amplitude of the signal template has been magnified to make its structure clear. The large excess power in the L1 data, not seen in the H1 data, despite comparable strain sensitivities and comparable sidereal-averaged antenna pattern sensitivities to any point in the sky, excludes an astrophysical source for the L1 artifact. The fact that the artifact aligns in frequency with the putative signal’s template peak in power confirms contamination of the outlier from an instrumental source. (color online).