pith. sign in

arxiv: 2606.07330 · v1 · pith:OVST64YPnew · submitted 2026-06-05 · 🌀 gr-qc · astro-ph.CO· astro-ph.HE· astro-ph.IM

Red noise and evolving signals: a complete frequentist approach to supermassive black hole binary searches with pulsar timing array

Xuan Tao , Boris Goncharov , Yiqian Qian , Yan Wang , Soumya D. Mohanty This is my paper

Pith reviewed 2026-06-27 21:15 UTC · model grok-4.3

classification 🌀 gr-qc astro-ph.COastro-ph.HEastro-ph.IM
keywords pulsar timing arraysgravitational wave detectionsupermassive black hole binariesfrequentist statisticsred noiseevolving signalslikelihood ratio test
0
0 comments X

The pith

A generalized likelihood ratio test enables reliable detection of evolving supermassive black hole binary signals in pulsar timing arrays that include red noise.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces a frequentist method to search for gravitational waves from evolving supermassive black hole binaries in pulsar timing array observations. Earlier frequentist approaches were limited to non-evolving sources and produced biased results when source evolution and unknown pulsar distances were considered alongside pulsar-specific red noise. The new approach uses a Generalized Likelihood Ratio Test and associated T-statistic that reduces the high-dimensional fitting problem through semi-analytical maximization over pulsar phase parameters followed by particle swarm optimization on the remaining parameters. Simulations for a chirp mass of 10^9.2 solar masses and signal-to-noise ratio of 20 demonstrate that this statistic reaches 100 percent detection probability at a false-alarm probability of 0.06 in a 30-pulsar array with 100 ns white noise and realistic red noise. This framework avoids the circular analysis of prior restricted methods and directly incorporates both Earth and pulsar terms in the signal model.

Core claim

The GLRT and T-statistic form a complete frequentist detection framework for isolated evolving SMBHBs in PTA data that simultaneously estimates signal and noise parameters, retains both Earth and pulsar terms, and reduces computational dimensionality via semi-analytical maximization over pulsar phases combined with particle swarm optimization to locate the global likelihood maximum.

What carries the argument

Generalized Likelihood Ratio Test (GLRT) and T-statistic using semi-analytical maximization over pulsar phase parameters followed by Particle Swarm Optimization over the remaining parameters.

If this is right

  • The statistic achieves 100 percent detection probability at false-alarm probability 0.06 for an evolving SMBHB with chirp mass 10^9.2 solar masses and SNR 20 in a 30-pulsar timing array with 100 ns rms white noise and pulsar-specific red noise.
  • The method incorporates both Earth and pulsar terms without forcing the circular analysis that biases earlier F-statistic approaches limited to non-evolving sources.
  • Dimensionality of the fit scales with the number of pulsars yet remains computationally tractable through the semi-analytical reduction step.
  • The framework directly models source evolution from gravitational-wave emission while estimating pulsar distances as unknown parameters.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Application to longer PTA data sets could allow population studies that include realistic orbital evolution rather than assuming fixed-frequency sources.
  • The same dimensionality-reduction technique might extend to searches for other time-varying gravitational-wave sources in PTA data.
  • Combining this frequentist statistic with Bayesian methods could provide cross-checks on detection significance when red noise properties are uncertain.

Load-bearing premise

The semi-analytical maximization of the likelihood over the pulsar phase parameters combined with Particle Swarm Optimization reliably locates the global maximum without missing superior solutions or introducing bias in the presence of red noise.

What would settle it

A set of Monte Carlo simulations on a 30-pulsar array with injected evolving SMBHB signals of chirp mass 10^9.2 solar masses, SNR 20, and realistic red noise in which the T-statistic fails to reach near-100 percent detection at a false-alarm probability of 0.06 would falsify the claimed performance.

Figures

Figures reproduced from arXiv: 2606.07330 by Boris Goncharov, Soumya D. Mohanty, Xuan Tao, Yan Wang, Yiqian Qian.

Figure 1
Figure 1. Figure 1: FIG. 1. Log-likelihood ln [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Time-domain representation of the simulated data [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Distributions of the detection statistic [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Comparison between the maximized log-likelihood [PITH_FULL_IMAGE:figures/full_fig_p010_4.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6. Comparison of CGW parameter estimation performance. Top panel [PITH_FULL_IMAGE:figures/full_fig_p013_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: FIG. 7. Time-domain reconstruction of the isolated CGW [PITH_FULL_IMAGE:figures/full_fig_p014_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: FIG. 8 [PITH_FULL_IMAGE:figures/full_fig_p014_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: FIG. 9. Recovered versus injected intrinsic red-noise param [PITH_FULL_IMAGE:figures/full_fig_p014_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: FIG. 10. Signal–noise leakage diagnostic for red-noise characterization. The plotted quantity is the recovered red-noise power [PITH_FULL_IMAGE:figures/full_fig_p015_10.png] view at source ↗
read the original abstract

Searches for gravitational waves (GWs) from isolated supermassive black hole binaries (SMBHBs) in pulsar timing array (PTA) data require simultaneous estimation of signal and noise parameters, so the dimensionality of the fit scales with the number of observed pulsars. This computational difficulty is exacerbated when source evolution from GW emission is included, since retaining both Earth and pulsar terms introduces the unknown pulsar distances. Existing frequentist methods such as the $\mathcal{F}$-statistic, restricted so far to non-evolving sources, effectively, imply a circular analysis, which may lead to biased estimators. We present a Generalized Likelihood Ratio Test (GLRT) and the associated $\mathcal{T}$-statistic that overcomes the aforementioned limitations. The formulation of the GLRT extends earlier work in which the dimensionality of the fitting problem was drastically reduced by semi-analytical maximization of the likelihood over the pulsar phase parameters, followed by efficient global optimization over the remaining parameters using Particle Swarm Optimization. Our simulations demonstrate that for an evolving SMBHB signal with chirp mass $\mathcal{M}=10^{9.2}\,M_\odot$ and signal-to-noise ratio $20$, this detection statistic achieves a $100\%$ detection probability at a false-alarm probability of $0.06$ in a 30-pulsar timing array, which is characterized by a $100~\mathrm{ns}$ root-mean-square white noise residual and pulsar-specific red noise.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 0 minor

Summary. The paper introduces a Generalized Likelihood Ratio Test (GLRT) and associated T-statistic for frequentist searches of evolving supermassive black hole binary (SMBHB) gravitational wave signals in pulsar timing array (PTA) data. It extends prior F-statistic work by using semi-analytical maximization of the likelihood over the 2N_pulsar phase parameters, followed by Particle Swarm Optimization (PSO) over the remaining parameters (including unknown pulsar distances), to avoid circularity and handle the increased dimensionality from source evolution. Simulations for a specific case (chirp mass 10^{9.2} M_odot, SNR 20) in a 30-pulsar array with 100 ns white noise plus pulsar-specific red noise report 100% detection probability at a false-alarm probability of 0.06.

Significance. If the optimization procedure is shown to reliably recover global likelihood maxima, the method would provide a computationally tractable frequentist framework for evolving SMBHB signals that addresses a key limitation of existing approaches. The dimensionality reduction via semi-analytical phase maximization is a clear technical strength, and the reported operating point (if robust) would be relevant for interpreting PTA data from arrays like NANOGrav or EPTA.

major comments (2)
  1. [Abstract] Abstract (simulation results paragraph): The headline performance (100% detection probability at FAP=0.06) is obtained by evaluating the GLRT/T-statistic at the numerically located maximum; however, no details are provided on the number of Monte Carlo realizations used to construct the empirical FAP distribution, the PSO convergence criteria, or any validation that single PSO runs recover the global maximum on the non-convex surface induced by red-noise covariance. This directly affects whether the reported detection probabilities are biased low.
  2. [Abstract] Abstract (method description): The claim that the semi-analytical maximization over pulsar phases combined with PSO 'overcomes the aforementioned limitations' of circular analysis requires explicit demonstration that the reduced-dimensional optimization does not introduce bias when pulsar distances are unknown and red noise is present; without such a test or derivation showing independence from fitted parameters, the independence from the F-statistic circularity cannot be confirmed.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their thorough review and valuable comments on our manuscript. We address each major comment below and have made revisions to the manuscript to incorporate the suggested improvements.

read point-by-point responses

  1. Referee: [Abstract] Abstract (simulation results paragraph): The headline performance (100% detection probability at FAP=0.06) is obtained by evaluating the GLRT/T-statistic at the numerically located maximum; however, no details are provided on the number of Monte Carlo realizations used to construct the empirical FAP distribution, the PSO convergence criteria, or any validation that single PSO runs recover the global maximum on the non-convex surface induced by red-noise covariance. This directly affects whether the reported detection probabilities are biased low.

    Authors: We agree that these methodological details were omitted from the abstract and main text. In the revised manuscript, we will expand the methods and simulation sections to report the number of Monte Carlo realizations used to construct the empirical FAP distribution, the PSO convergence criteria, and validation results from multiple PSO runs with varied initializations to confirm reliable recovery of the global maximum. These additions will allow assessment of whether the reported detection probabilities are robust. revision: yes

  2. Referee: [Abstract] Abstract (method description): The claim that the semi-analytical maximization over pulsar phases combined with PSO 'overcomes the aforementioned limitations' of circular analysis requires explicit demonstration that the reduced-dimensional optimization does not introduce bias when pulsar distances are unknown and red noise is present; without such a test or derivation showing independence from fitted parameters, the independence from the F-statistic circularity cannot be confirmed.

    Authors: The semi-analytical maximization over the 2N_pulsar phase parameters is derived to be independent of the remaining parameters (including distances) by construction, as the likelihood is maximized analytically for any fixed values of the other parameters before PSO is applied. To explicitly demonstrate the absence of bias under unknown distances and red noise, the revised manuscript will include additional simulations comparing the full procedure against cases with fixed distances and against the non-evolving limit, confirming consistency and lack of introduced bias. revision: yes

Circularity Check

0 steps flagged

No significant circularity; performance claims rest on independent Monte Carlo simulations

full rationale

The paper defines a GLRT/T-statistic via semi-analytical maximization over pulsar phases plus PSO on remaining parameters, then reports empirical detection probabilities from separate Monte Carlo trials on synthetic PTA data with injected signals and red noise. These operating-point numbers are not obtained by fitting to the same data used for the statistic definition, nor do they reduce to a self-citation chain or ansatz smuggled from prior work. The reference to 'earlier work' is only for the maximization technique; the present results are self-contained against external simulation benchmarks. No load-bearing step equates a claimed prediction to its own inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Only the abstract is available, so no specific free parameters, axioms, or invented entities can be extracted from the full text; the central claim rests on the validity of the dimensionality reduction technique and the simulation setup described at high level.

pith-pipeline@v0.9.1-grok · 5818 in / 1140 out tokens · 18135 ms · 2026-06-27T21:15:41.917677+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

51 extracted references · 3 linked inside Pith

  1. [1]

    let-them-fly

    were later adapted to PTAs, most notably as the Fe-statistic and theF p-statistic [24–26]. These statis- tics are attractive because they are easy to implement and, under fixed-noise assumptions, lead to simple ref- erence distributions for threshold setting [27]. However, this simplicity comes from approximations. The coherent Fe-statistic omits the puls...

    Pith/arXiv arXiv 2026

  2. [2]

    CGW Parameter Recovery We define the estimation error for a parameterλas ∆λ=λ best −λ true. To visualize the estimation accu- racy while accounting for scale variations across differ- ent signal strengths, Figure 6 displays the kernel den- sity estimates (KDE) of the normalized estimation errors (∆λ/σ∆λ) for both the 20-pulsar baseline (panel a) and the 3...

  3. [3]

    Red Noise Characterization A central motivation for the joint-search framework is to characterize the covariance between a deterministic continuous gravitational wave (CGW) signal and intrin- sic pulsar red noise within the same likelihood analysis. Even when red-noise and CGW parameters are sampled simultaneously in a Bayesian stage, a subsequent frequen...

  4. [4]

    TheH 1-fit covariance result provides a useful diagnos- tic of signal–noise covariance

    This slightly exceeds theH 0-covarianceF p result at the weakest injection, consistent with the expectation that separately profiling the noise parameters underH 0 andH 1 can reduce weak-signal loss from a noise-only covariance estimate. TheH 1-fit covariance result provides a useful diagnos- tic of signal–noise covariance. For signal-containing data, it ...

    2000

  5. [5]

    M. V. Sazhin, Soviet Astronomy22, 36 (1978)

    1978

  6. [6]

    Detweiler, Astrophys

    S. Detweiler, Astrophys. J.234, 1100 (1979)

    1979

  7. [7]

    R. S. Foster and D. C. Backer, Astrophys. J.361, 300 (1990)

    1990

  8. [8]

    Hobbset al., Class

    G. Hobbset al., Class. Quantum Grav.27, 084013 (2010)

    2010

  9. [9]

    Burke-Spolaoret al., Astron

    S. Burke-Spolaoret al., Astron. Astrophys. Rev.27, 5 (2019)

    2019

  10. [10]

    Agazieet al.(NANOGrav Collaboration), Astrophys

    G. Agazieet al.(NANOGrav Collaboration), Astrophys. J. Lett.951, L8 (2023)

    2023

  11. [11]

    Antoniadiset al.(EPTA Collaboration and InPTA Collaboration), Astron

    J. Antoniadiset al.(EPTA Collaboration and InPTA Collaboration), Astron. Astrophys.678, A3 (2023)

    2023

  12. [12]

    D. J. Reardonet al.(PPTA Collaboration), Astrophys. J. Lett.951, L6 (2023)

    2023

  13. [13]

    Xuet al.(CPTA Collaboration), Res

    H. Xuet al.(CPTA Collaboration), Res. Astron. Astro- phys.23, 075024 (2023)

    2023

  14. [14]

    Tarafdaret al., Publ

    P. Tarafdaret al., Publ. Astron. Soc. Aust.39, e053 (2022)

    2022

  15. [15]

    R. W. Hellings and G. S. Downs, Astrophys. J.265, L39 (1983). 17

    1983

  16. [16]

    Rajagopal and R

    M. Rajagopal and R. W. Romani, Astrophys. J.446, 543 (1995)

    1995

  17. [17]

    A. H. Jaffe and D. C. Backer, Astrophys. J.583, 616 (2003)

    2003

  18. [18]

    Sesana, A

    A. Sesana, A. Vecchio, and C. N. Colacino, Mon. Not. R. Astron. Soc.390, 192 (2008)

    2008

  19. [19]

    L. Z. Kelley, L. Blecha, and L. Hernquist, Mon. Not. R. Astron. Soc.464, 3131 (2017)

    2017

  20. [20]

    C. M. F. Mingarelli, K. Grover, T. Sidery, R. J. E. Smith, and A. Vecchio, Phys. Rev. Lett.109, 081104 (2012)

    2012

  21. [21]

    Y. Wang, S. D. Mohanty, and Z. Cao, The Astrophys- ical Journal Letters907, L43 (2021), arXiv:2012.04261 [astro-ph.HE]

    arXiv 2021

  22. [22]

    X. Tao, Y. Wang, and S. D. Mohanty, Phys. Rev. D111, 083030 (2025), arXiv:2412.07615 [astro-ph.HE]

    arXiv 2025

  23. [23]

    Corbin and N

    V. Corbin and N. J. Cornish, arXiv e-prints (2010), arXiv:1008.1782 [astro-ph.GA]

    Pith/arXiv arXiv 2010

  24. [24]

    K. J. Leeet al., Mon. Not. R. Astron. Soc.414, 3251 (2011)

    2011

  25. [25]

    A. T. Delleret al., Astrophys. J. Lett.685, L67 (2008)

    2008

  26. [26]

    J. P. W. Verbiestet al., Astrophys. J.755, 39 (2012)

    2012

  27. [27]

    Jaranowski, A

    P. Jaranowski, A. Kr´ olak, and B. F. Schutz, Phys. Rev. D58, 063001 (1998)

    1998

  28. [28]

    Babak and A

    S. Babak and A. Sesana, Phys. Rev. D85, 044034 (2012)

    2012

  29. [29]

    J. A. Ellis, X. Siemens, and J. D. E. Creighton, Astro- phys. J.756, 175 (2012)

    2012

  30. [30]

    Zhuet al., Mon

    X. Zhuet al., Mon. Not. R. Astron. Soc.444, 3709 (2014)

    2014

  31. [31]

    Arzoumanianet al., Astrophys

    Z. Arzoumanianet al., Astrophys. J.794, 141 (2014)

    2014

  32. [32]

    van Haasteren and M

    R. van Haasteren and M. Vallisneri, Phys. Rev. D90, 104012 (2014)

    2014

  33. [33]

    J. A. Elliset al., Enterprise: Enhanced numerical toolkit evaluator for robust pulsar inference supported by em- pirical data (2019), astrophysics Source Code Library, record ascl:1912.015

    2019

  34. [34]

    S. J. Vigeland, K. Islo, S. R. Taylor, and X. Siemens, Phys. Rev. D98, 044003 (2018)

    2018

  35. [35]

    Aggarwal,et al

    NANOGrav Collaboration, K. Aggarwal,et al. (NANOGrav Collaboration), Astrophys. J.880, 116 (2019)

    2019

  36. [36]

    Arzoumanianet al.(NANOGrav Collaboration), As- trophys

    Z. Arzoumanianet al.(NANOGrav Collaboration), As- trophys. J. Lett.951, L50 (2023)

    2023

  37. [37]

    B´ ecsy, N

    B. B´ ecsy, N. J. Cornish, and L. Z. Kelley, Astrophys. J. 931, 95 (2022)

    2022

  38. [38]

    S. C. Sardesai, G. E. Freedman, S. J. Vigeland, and C. A. Witt, Phys. Rev. D113, 084067 (2026)

    2026

  39. [39]

    Kennedy and R

    J. Kennedy and R. Eberhart, inProceedings of ICNN’95- International Conference on Neural Networks(1995) pp. 1942–1948

    1995

  40. [40]

    Mohanty,Swarm intelligence methods for statistical regression(CRC Press, 2018)

    S. Mohanty,Swarm intelligence methods for statistical regression(CRC Press, 2018)

    2018

  41. [41]

    Lentati, P

    L. Lentati, P. Alexander, M. P. Hobson, S. Taylor, J. Gair, S. T. Balan, and R. van Haasteren, Phys. Rev. D87, 104021 (2013)

    2013

  42. [42]

    Y. Wang, S. D. Mohanty, and F. A. Jenet, Astrophys. J. 815, 125 (2015)

    2015

  43. [43]

    Wang and S

    Y. Wang and S. D. Mohanty, Phys. Rev. Lett.118, 151104 (2017), arXiv:1611.09440 [astro-ph.IM]

    Pith/arXiv arXiv 2017

  44. [44]

    Y.-Q. Qian, S. D. Mohanty, and Y. Wang, Phys. Rev. D 106, 023016 (2022)

    2022

  45. [45]

    Wang and S

    Y. Wang and S. D. Mohanty, Phys. Rev. D81, 063002 (2010)

    2010

  46. [46]

    T. S. Weerathunga and S. D. Mohanty, Phys. Rev. D95, 124030 (2017)

    2017

  47. [47]

    Zhanget al., Phys

    X.-H. Zhanget al., Phys. Rev. D104, 024023 (2021)

    2021

  48. [48]

    Y. Fu, Y. Wang, and S. D. Mohanty, Phys. Rev. D111, 043026 (2025), arXiv:2407.10797 [astro-ph.IM]

    arXiv 2025

  49. [49]

    S. M. Kay,Fundamentals of Statistical Signal Processing, Volume II: Detection Theory(Prentice Hall, 1998)

    1998

  50. [50]

    Neyman and E

    J. Neyman and E. S. Pearson, Philos. Trans. R. Soc. Lon- don A231, 289 (1933)

    1933

  51. [51]

    S. R. Taylor, P. T. Baker, J. S. Hazboun, J. Simon, and S. J. Vigeland, enterprise extensions (2021), v2.4.3. Appendix A: Simulated Pulsar Parameters and Signal-to-Noise Ratios This appendix details the noise properties and signal contributions for the simulated PTAs. Table V lists the injected red noise parameters (log 10 ARN andγ RN), dis- tances (L I),...

    2021