⚛️ general relativity

Orientation-dependent sensitivity of the all-sky all-frequency gravitational-wave radiometer search to continuous-wave sources

This paper demonstrates that the all-sky all-frequency gravitational-wave radiometer search systematically underestimates the strength of continuous-wave sources due to orientation-dependent polarization effects, deriving a universal average projection factor of $2/5$ and showing that this necessitates significantly higher upper limits on strain amplitude for sub-threshold candidates.

Original authors: Stavros Venikoudis, Giacomo Bruno

Published 2026-10-22

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Stavros Venikoudis, Giacomo Bruno

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Gravitational waves are ripples in the fabric of space-time, created when massive objects like neutron stars or black holes accelerate. When these waves reach Earth, they stretch and squeeze the distance between objects by a tiny amount. To catch these faint signals, scientists use detectors that act like giant ears, listening for the specific hum of a spinning, deformed neutron star. These stars, if they have a slight bump on their surface, emit a steady, continuous tone as they rotate. Finding them is like trying to hear a single violin playing a constant note in a stadium full of noise. The challenge is not just finding the note, but understanding how loud the violin actually is, because the sound we hear depends entirely on how the instrument is tilted relative to our ears.

A recent study by Stavros Venikoudis and Giacomo Bruno at the University of Louvain in Belgium addresses a subtle but critical flaw in how scientists have been interpreting these signals. For years, a specific method called the "all-sky all-frequency radiometer" has been used to scan the entire universe for these continuous waves. This method works by comparing data from two widely separated detectors to find a signal that appears in both. However, the method was originally designed to listen for a background hum made of many overlapping sources, which is expected to be unoriented and random. When the researchers applied this same method to look for individual, spinning neutron stars, they found it was systematically underestimating the true strength of the source. The method assumes the signal is equally strong in all directions of polarization, but a spinning star emits sound differently depending on whether we see it face-on or from the side.

The researchers discovered that this assumption creates a bias. If a neutron star is spinning directly toward us, the method measures its full power. But if the star is spinning on its side, or at any angle in between, the method records less power than is actually there. It is as if a lighthouse beam is brightest when pointed directly at a ship, but dimmer when seen from the side; the radiometer method, however, assumes the light is always the same brightness regardless of the angle. This leads to a significant error: a very strong star spinning on its side might look so faint in the data that it falls below the threshold for detection, leading scientists to believe it does not exist or is much weaker than it truly is.

To fix this, the team developed a way to calculate a "projection factor," a number that describes exactly how much power is lost based on the star's tilt. They proved that if you average this factor over all possible angles, the method captures only 40 percent of the true signal power on average. This means that for a typical, randomly oriented star, the method underestimates its strength by a factor of roughly 1.6. The researchers tested this new understanding using four simulated signals that were physically injected into the LIGO detectors during a previous observing run. These simulations had known strengths and known angles. When they applied their correction, they found that a signal which appeared too weak to be detected could, in reality, be a powerful source that was simply viewed from an unfavorable angle. One such signal, which looked like a weak whisper to the uncorrected method, was revealed to be a strong shout when the angle was accounted for.

The study then turned to real data from the most recent observing run, which included 505 candidate signals that were just below the detection threshold. The researchers showed that for each of these candidates, the reported limit on the star's strength was not a single number, but a range. The value published in previous reports assumed the star was facing us directly, which is the most favorable case. However, if the star is tilted, the true limit on its strength could be nearly three times higher. This means that a candidate signal that looks like a faint possibility might actually be a very strong source that is just tilted away from us. The researchers calculated that for these candidates to be detected, the stars would likely need to be tilted less than 35 to 50 degrees from a face-on position. Since stars are randomly oriented in space, this means only about 18 to 36 percent of such strong stars would be visible to the current search method.

The implications of this work are practical and immediate. The significance of the candidate signals—the statistical proof that they are real and not just noise—does not change. A signal that was just below the limit is still just below the limit. However, the interpretation of what that signal means changes completely. A low measurement no longer implies a weak star; it could be a very strong star seen from the side. The authors recommend that future searches report a range of possible strengths rather than a single value, and that they consider the orientation of the source when deciding which signals to investigate further. By accounting for the tilt of the cosmic violins, scientists can stop missing the loudest players in the orchestra simply because they are standing at an angle.

Technical Summary: Orientation-Dependent Sensitivity of the ASAF Gravitational-Wave Radiometer Search

Problem Statement
The all-sky all-frequency (ASAF) gravitational-wave radiometer search is designed to detect persistent, narrowband signals, such as continuous waves (CW) from deformed neutron stars, by cross-correlating data from pairs of detectors. The standard ASAF statistic assigns equal weight to the $+$ and $\times$ tensor polarizations. While this weighting is optimal for an unpolarized stochastic gravitational-wave background (SGWB), it is suboptimal for CW sources. CW sources emit with polarization amplitudes determined by their inclination angle $\iota$ and polarization angle $\psi$ . Consequently, unless a source is viewed face-on ( $\iota=0$ ), the equal-weight statistic measures less power than a face-on source of the same intrinsic amplitude would produce. This leads to an "orientation bias" where the search underestimates the intrinsic strength of the source and reduces the signal-to-noise ratio (SNR), potentially causing real signals to fall below detection thresholds.

Methodology
The authors derive the exact dependence of this bias on source orientation and propagate it through the ASAF analysis pipeline.

Derivation of the Projection Factor ( $\mu^2$ ):
The paper defines a projection factor $\mu^2(\iota, \psi)$ as the ratio of the effective strain power measured by the ASAF statistic to the intrinsic strain power ( $h_0^2$ ) of a CW source.
- The authors derive a closed-form expression for $\mu^2(\iota, \psi)$ (Eq. 33), showing it consists of a term dependent only on inclination and a term modulating with the polarization angle $\psi$ as a single harmonic ( $4\psi$ ).
- The modulation depth $\epsilon$ is determined by the baseline geometry and the source's declination.
- The authors prove that the isotropic average of this factor over all orientations is exactly $2/5$ (0.4), independent of sky position, baseline geometry, noise weighting, or frequency.
Propagation to Search Quantities:
The formalism is applied to the key outputs of the ASAF search:
- Strain Amplitude: The intrinsic amplitude $h_0$ is related to the reported effective strain $h_{\text{eff}}$ by $h_0 = h_{\text{eff}} / \sqrt{\mu^2}$ .
- Signal-to-Noise Ratio (SNR): The recovered SNR for a source at a specific orientation is $\rho_{\text{rec}} = \mu^2 \rho_{\text{ref}}$ , where $\rho_{\text{ref}}$ is the SNR of a face-on source with the same intrinsic strength.
- Upper Limits (ULs): Bayesian upper limits on $h_0$ are derived by marginalizing the likelihood over the unknown orientation parameters ( $\iota, \psi$ ), replacing the standard assumption of circular polarization ( $\mu^2=1$ ) with the distribution of $\mu^2$ .
Validation and Application:
- O2 Hardware Injections: The formalism is tested against four simulated pulsar signals (hardware injections) from the second observing run (O2) with known parameters and varying orientations (from nearly face-on to nearly edge-on).
- O4a Candidates: The method is applied to 505 sub-threshold candidates selected from the first part of the fourth observing run (O4a) for follow-up investigation.

Key Results

Analytical Proof: The paper confirms that the average projection factor is exactly $2/5$ for the full ASAF statistic at any sky position. It also provides the exact closed form for the residual dependence on $\psi$ , identifying the least favorable orientations and the maximum detectable inclination for any given sky location.
Hardware Injection Analysis:
- For a nearly edge-on injection ( $\iota \approx 94.6^\circ$ ), the measured SNR was 5.20, below the detection threshold of 6.19. However, correcting for the projection factor ( $\mu^2 \approx 0.167$ ) reveals that the same signal, if viewed face-on, would have an SNR of $31.1 \pm 6.0$ .
- Bayesian upper limits on the intrinsic strain amplitude $h_0$ , marginalized over orientation, were found to be 2.7 to 2.9 times larger than the standard face-on (circular polarization) limits.
- The analysis successfully recovered the injected amplitudes when correcting for $\mu^2$ and accounting for known spectral losses (Doppler and pixel-phase effects).
O4a Candidate Analysis:
- For the 505 sub-threshold candidates, the marginalized upper limits increase by a factor of 2.5 to 2.9 compared to the published circular limits.
- If a candidate is a real CW source, it would only reach the detection threshold if its inclination is less than $34.7^\circ$ to $50.3^\circ$ from face-on (depending on the specific candidate's SNR). This corresponds to an orientation coverage of only 18% to 36% of the isotropic sky.
- The "worst-case" upper limits (assuming the least favorable orientation) can be up to $\sim 3.6$ times the circular limit.

Significance and Claims
The paper claims that accounting for the orientation bias $\mu^2$ fundamentally changes the interpretation of ASAF results regarding continuous-wave sources, without altering the statistical significance of the candidates.

Reinterpretation of Non-Detections: A low measured SNR does not necessarily imply a weak intrinsic source; it may indicate a strong source at an unfavorable orientation. The intrinsic strength behind a given SNR can be up to $1/\mu^2_{\text{min}} \approx 8$ to 13 times larger than the face-on estimate.
Reporting Standards: The authors recommend that future ASAF analyses report upper limits as a range over orientation (or a marginalized value) rather than a single face-on value. They also suggest reporting the maximum detectable inclination ( $\iota_{\text{max}}$ ) and orientation coverage ( $C_{\text{orient}}$ ) alongside the measured SNR and effective strain.
Selection Bias: Selecting candidates based solely on measured SNR introduces a bias toward face-on sources. Follow-up searches should account for the fact that sub-threshold candidates could be strong sources at high inclinations.

The paper concludes that while the measured SNR and p-values of candidates remain unchanged, the physical interpretation of the strain amplitude and detection probability requires the inclusion of the orientation-dependent projection factor.

Technical Summary: Orientation-Dependent Sensitivity of the ASAF Gravitational-Wave Radiometer Search

More like this