{"id":"f6e25a54-0437-423b-9d11-a2e73a585862","arxiv_id":"2412.01899","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Time-dependent fluctuations in the gravitational wave background can be large and detectable when the background is dominated by massive, eccentric supermassive black hole binaries or by a single eccentric binary with signal-to-noise ratio near one.","lead":"placeholder","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Detectability claim rests on an unvalidated ΔΩ²-vs-σ²S comparison; no detection statistic or end-to-end simulation is provided, so the central claim is conditional at best.","rationale":"The paper makes a credible new analytical derivation: the variance of the GW luminosity of an eccentric binary (Eq. 4) and the covariance of harmonics (Eq. 10) are internally consistent, verified numerically to 10^-14 (Fig. 10). The 1/N scaling for equal populations (Eq. 30) is sound. The phase/frequency evolution assumption flagged by the reader is not the main risk because secular evolution is negligible in the PTA band. The actual load-bearing weakness is the step from calculated ΔΩ² to the detectability conclusion. The comparison to σ²S mixes a per-frequency signal variance with a broadband amplitude uncertainty, and no detection statistic is defined. The paper's own text flags this ('beyond the scope of this article', 'realistic simulations are essential'), so the central claim should be treated as conditional on a future end-to-end detection study. The reader's verdict of CONDITIONAL is appropriate. The population parameters for POP A/B are also only partially described, but the missing detection study is the more decisive issue.","tokens_in":16427,"tokens_out":12430,"duration_ms":136817,"concrete_test":"Perform an end-to-end PTA simulation: generate 10 years of synthetic TOAs for an EPTA-like array (25 pulsars, ~3-day cadence, realistic red and DM noise), inject the nonstationary GWB predicted for POP B (same population parameters, ΔΩ² up to ~10² at relevant frequencies), and run a Bayesian analysis comparing a stationary GWB model with a nonstationary model that includes a time-varying variance parameter. If the injected ΔΩ² is not recovered with a decisive Bayes factor (e.g., ln B > 5) at the frequencies where ΔΩ² > σ²S, the detectability claim in the paper fails. A cheaper auxiliary check: verify that the per-frequency uncertainty σ²S(f) (not the broadband amplitude uncertainty) is actually larger than ΔΩ²(f) before claiming detectability.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is that realistic eccentric SMBHB populations, or a single low-SNR eccentric binary, produce nonstationary GWB fluctuations detectable by current/future PTAs. The evidence for detectability is a comparison (Sec VI) between ΔΩ²(f) and σ²S, the Fisher-forecast relative uncertainty on the GWB amplitude A² under stationary Gaussian noise. This is not a detection statistic: σ²S is a broadband constraint on a power-law amplitude, not the per-frequency variance of a spectrum estimator, and the paper never constructs a likelihood or model comparison that would detect the time-varying variance. The authors themselves acknowledge this in Sec VI A ('A more in-depth analysis would require defining detection statistics') and Sec VI B ('realistic simulations are essential'). Without such a statistic, the statement that ΔΩ² > σ²S implies detectability is unsupported. The reader's flagged phase-evolution assumption, by contrast, is well justified: for SMBHBs in the PTA band, the chirp timescale is ≳10^8 yr, so Φ(t) ≈ 2πfp(t−t0) over a 10-yr observation is accurate to better than 10^-6, and the time-average/ensemble-average mapping is valid for uniformly distributed phases. The load-bearing gap is therefore the missing detection study, not the phase mapping.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies the time-dependent (nonstationary) character of the gravitational wave background (GWB) produced by eccentric supermassive black hole binaries (SMBHBs) in the pulsar timing array (PTA) band. The authors first derive closed-form expressions for the variance of the GW luminosity of a single eccentric binary within one orbital period, and for the covariance between the harmonic components of the luminosity, starting from the Peters-Mathews quadrupole formulas and a Fourier decomposition of the Keplerian orbit. They verify the consistency of these expressions numerically. They then compute the normalized variance ΔΩ²(f) of the GWB power spectral density for three scenarios: a toy population of equal, homogeneously distributed binaries; two realistic astrophysical population models (POP A and POP B) built from merger-rate and SMBH-host prescriptions; and a single eccentric binary superposed on a stationary GWB. Finally, they compare ΔΩ² with the Fisher-forecast relative uncertainty σ²_S on the GWB amplitude for three PTA configurations (EPTA10, IPTA20, SKA10) and conclude that realistic massive eccentric populations, or even individual binaries with SNR ≈ 1, could produce detectable nonstationary fluctuations.","tokens_in":16789,"tokens_out":5199,"duration_ms":59250,"significance":"The analytical derivation of the luminosity variance and harmonic covariance (Eqs. 4, 10, and Appendix A) is a useful and self-contained contribution, with the internal consistency check in Fig. 10 providing strong support for the algebra. The finding that a single low-SNR eccentric binary can produce relative power fluctuations of order unity, while deterministic searches would struggle to detect it, is interesting and could motivate new data-analysis strategies. The neglect of binary frequency evolution (Sec. II A 3 and Appendix A 2 a) is well justified for SMBHBs in the PTA band, since the chirp timescale vastly exceeds any realistic observation span; I do not regard that assumption as a substantive weakness. However, the paper's central detectability claim is currently supported only by a heuristic comparison between ΔΩ² and a broadband Fisher uncertainty, not by an actual detection statistic or end-to-end simulation. The authors explicitly acknowledge this gap, but the abstract and conclusions present the nonstationarity as 'detectable' without qualification.","major_comments":[{"comment":"The comparison ΔΩ²(f) > σ²_S is not a valid detection criterion. σ²_S is the Fisher-forecast relative uncertainty on the broadband GWB amplitude A² under the assumption of a stationary, Gaussian GWB with fixed spectral index; it is not the per-frequency variance of the PSD estimator S(f). Table I lists single numbers for each PTA, whereas ΔΩ²(f) varies strongly with frequency. The paper does not construct a likelihood, a model comparison, or a statistic that would actually measure time variations of the PSD, and the authors acknowledge this in Sec. VI A ('A more in-depth analysis would require defining detection statistics'). As it stands, the abstract's statement that the nonstationarity 'might become very large and detectable' and the conclusion's 'easily detectable' are not supported by the evidence presented. This is the load-bearing weakness: either a proper detection study (or at least a per-frequency variance estimate of the spectrum estimator) must be provided, or the detectability claims must be substantially softened.","section":"Sec. VI, Eqs. (38) and Table I"},{"comment":"The population-simulation estimates of ΔΩ²(f) are obtained from a finite number of sampled binaries ('hundreds of thousands'), but the paper does not state the total number of SMBHBs represented by the sampling nor demonstrate convergence of ΔΩ² with respect to that total number. Since ΔΩ² scales as 1/N for a homogeneous population (Eq. 30), and since the bright, massive binaries dominate the variance, the median and percentile bands in Fig. 6 could depend on the sampling cutoff or the realization size. Please specify the total number of binaries in each simulated universe, the effective number of sources contributing to each frequency bin, and show that the results are converged with respect to the sample size.","section":"Sec. IV B and Fig. 6"},{"comment":"The approximation cov{Pn,Pm} ≈ ⟨Pn⟩⟨Pm⟩[(G/F)²-1] does not follow from the equality of weighted sums in Eq. (A18). From Σ_{n,m} g(n,e)g(m,e)[(G/F)²-1] = Σ_{n,m} g(n,e)g(m,e)(⟨fn fm⟩-1), one can only conclude that the weighted average of (⟨fn fm⟩-1) equals (G/F)²-1, not that every individual term satisfies the relation. The derivation presented in Eqs. (A19)-(A23) therefore overstates the rigor of the approximation. Since the main single-source results (Figs. 7 and 8) use the exact covariance from Eq. (10) and only Fig. 9 uses the approximate formula, this issue is not fatal, but the approximation should be presented as a heuristic and its range of validity tested against the exact calculation.","section":"Appendix A 3, Eqs. (A18)-(A23)"},{"comment":"Equation (23) contains an unclear summation structure: the expression 'X_{i=1}' has no upper limit, and the subsequent sum over harmonics 'X_n' is not clearly nested. As written, it is not possible to determine whether the sum over i runs over the N(z)p(f_p,k) sources in the frequency bin or is a typographical artifact. Please clarify the indexing and the definition of Θ(f) so that Eqs. (26)-(27) follow unambiguously.","section":"Sec. III A 1, Eq. (23)"}],"minor_comments":[{"comment":"σ²_S(f) is written as a function of frequency, but the Fisher forecast in [9] yields a single relative uncertainty on A². Please clarify the frequency dependence or define σ²_S as a scalar.","section":"Eq. (38) and Table I"},{"comment":"The caption reads 'Cross-correlated fluctuations ΔΩ²(f, f′' with a missing closing parenthesis; it should read ΔΩ²(f, f′).","section":"Fig. 4 caption"},{"comment":"The notation 'NP' is used inconsistently: the text writes 'N (z)p(fp,k)' but the equation uses 'N (z)p(fp,k)' with an implicit product. The summation limits and the role of Θ(f) should be stated more explicitly.","section":"Sec. III A 1, just above Eq. (23)"},{"comment":"The name 'Peter and Mathews' appears in the appendix text while the reference is 'Peters and Mathews'; please correct the spelling.","section":"Appendix A, first paragraph and Eq. (A1)"},{"comment":"The labels 'EPT A10', 'IPT A20', 'SKA10' appear with an extra space in 'EPT A10'; please standardize the notation.","section":"Sec. VI B"},{"comment":"The sentence 'The contribution of consecutive higher harmonics increases 'in phase' as the binary accelerates' is somewhat vague; a more precise statement would be that the fn(t,e) are strongly correlated for nearby n, as shown in Fig. 2.","section":"Sec. II A 4, Eq. (11)"}],"recommendation":"major_revision","confidential_remarks":"The paper fits the scope of the journal and the core variance/covariance derivation is solid and internally checked. The main concern for publication is the gap between the computed ΔΩ² and the claimed detectability: the comparison with a broadband Fisher variance is not a detection statistic. I would suggest the authors either (a) add an end-to-end simulation or a proper likelihood-based test for nonstationarity, or (b) substantially temper the abstract and conclusions to state that the fluctuations are 'potentially observable' or 'could be detectable' pending a dedicated search. The population-simulation convergence issue (total number of binaries, sampling completeness) should also be addressed. With those changes, the paper would be a useful contribution to the GWB and PTA literature."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new thing here is the closed-form variance of the GW luminosity for an eccentric binary, Equation 4, plus the covariance structure between harmonics. That derivation is clean, self-contained, and they check it against the harmonic decomposition to relative error 10^-14. That part is good, real work. The application to nonstationarity of the GWB—showing that fluctuations scale as 1/N and that a realistic population or a single low-SNR eccentric binary can produce large ΔΩ²—is also new and worth taking seriously. The internal arithmetic and the citation pattern look solid; the population models are drawn from established sources, and the authors are honest about the leading-order post-Newtonian limitation.\n\nThe soft spot is the detectability argument in Section VI. They compare ΔΩ²(f) to σ²S, the Fisher-forecast relative uncertainty on the GWB amplitude A², and conclude that if ΔΩ² > σ²S the nonstationarity is detectable. But σ²S is a broadband constraint on a power-law amplitude under stationary Gaussian noise; it is not the per-frequency variance of a spectral estimator, and it is not a detection statistic for time-varying variance. The authors themselves say 'A more in-depth analysis would require defining detection statistics,' which is exactly right, but it means the central claim—that current and future PTAs can detect these fluctuations—is not yet supported. That is an addressable gap, not a fatal one: the paper provides the physical amplitude of the effect, and a proper likelihood or simulation-based study could settle it. I would not desk-reject on this; I would send it to a referee with the explicit request that the detectability section be rewritten around a defined statistic.\n\nThe phase-evolution concern raised in the report is minor. Over a 10-year observation, binaries in the PTA band have chirp timescales far longer than the observation span, so Φ(t) ≈ 2πfp(t−t0) is accurate and the time-average-to-ensemble-average mapping is fine. The real burden is the missing detection study.\n\nWho benefits: PTA data analysts thinking about non-Gaussian or nonstationary backgrounds, and people modeling eccentric SMBHB populations. It deserves a serious referee, but the verdict should be conditional on a proper detection analysis and on sharing the population simulation parameters. I would bring it to reading group to discuss exactly where the detectability argument breaks, and I would probably cite Equation 4 in my own work on eccentric backgrounds.","headline":"Solid new analytic result on luminosity variance for eccentric binaries, but the PTA detectability claim compares the wrong quantities and needs a real detection study before it can stand.","tokens_in":697,"tokens_out":983,"would_cite":true,"duration_ms":26039,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["04.30.-w","95.85.Sz","04.80.Nn"],"model":"deepseek-v4-flash","headline":"Eccentric supermassive black hole binaries can make the gravitational-wave background detectably flicker.","keywords":["gravitational wave background","eccentric binaries","supermassive black hole binaries","pulsar timing arrays","nonstationarity","spectral fluctuations","GW luminosity variance","nanohertz gravitational waves"],"falsifier":"Compare the measured variance of the GWB spectrum across independent time segments of a high-cadence pulsar timing array dataset with the variance expected from stationary noise: a source with signal-to-noise ratio $\\rho \\approx 1.4$ and eccentricity $e=0.5$ should produce an excess fluctuation $\\Delta\\Omega^2$ of order one localized at its harmonic frequencies, and a realistic massive eccentric population should produce $\\Delta\\Omega^2$ of order $10^2$; the absence of such excess variance at the predicted level would contradict the paper's central claim.","tokens_in":16201,"feed_emoji":"📡","tokens_out":15681,"duration_ms":127570,"temperature":0.7,"pith_summary":"The paper asks whether the stochastic gravitational wave background (GWB) produced by a population of massive black hole binaries is stationary, and shows that it need not be. Because an eccentric binary radiates most strongly near periastron, its GW power depends on its instantaneous orbital phase, so a population of such binaries has a spectrum that fluctuates in time. The authors quantify these fluctuations by $\\Delta\\Omega^2$, the variance of the spectrum divided by its mean squared, and find that for a homogeneous population of equal binaries the fluctuations are tiny, scaling as $1/N$ with the number of sources. For astrophysically realistic populations, where a small number of massive, highly eccentric binaries dominate the signal, the fluctuations can reach $\\Delta\\Omega^2 \\sim 10^2$, far above the relative measurement uncertainty $\\sigma_S^2 \\approx 0.9\\text{--}8\\times10^{-3}$ of current and planned pulsar timing arrays. Even a single eccentric binary with signal-to-noise ratio near unity, invisible to standard deterministic searches, can produce order-unity fluctuations, so detecting nonstationarity would reveal bright eccentric sources in the universe.","feed_headline":"Eccentric black hole binaries make the GW background flicker","feed_subtitle":"Even a single faint eccentric binary could leave a detectable flicker in the gravitational-wave background.","key_machinery":"The central object is the variance–covariance structure of the harmonic decomposition of the GW luminosity. Writing each harmonic as $P_n=\\langle P_n\\rangle f_n(t,e)$, the covariance between two harmonics of the same binary is $\\mathrm{cov}\\{P_n,P_m\\}=P_0^2 g(n,e)g(m,e)(\\langle f_n f_m\\rangle-1)$, which encodes the fact that harmonics fluctuate together as the binary accelerates through periastron. From this the paper forms the time-fluctuation statistic $\\Delta\\Omega^2(f)=\\mathrm{var}\\{h_c^2(f)\\}/\\langle h_c^2(f)\\rangle^2$, and for a single source derives the practical approximation $\\Delta\\Omega^2 \\approx \\rho^4[(G(e)/F(e))^2-1]$, where $F(e) = (1 + 73e^2/24 + 37e^4/96)/(1-e^2)^{7/2}$ amplifies the mean luminosity and $G(e)$ is the analogous factor for the squared luminosity. The machinery carries the argument by turning the burst-like radiation of eccentric binaries into the relative variance of the spectrum, a number that can be compared directly with pulsar timing array measurement uncertainties.","core_discovery":"The central claim is that the time variability of the gravitational wave background from supermassive black hole binaries is controlled by the variance of the GW luminosity of an eccentric binary within one orbital period, and that in realistic populations this variance is not averaged away. The paper computes the luminosity variance from the harmonic decomposition of the quadrupole moment and finds it scales as $P_0^2[G^2(e)-F^2(e)]$, where $F(e)$ amplifies the mean luminosity and $G(e)$ is the corresponding factor for the squared luminosity; this quantity vanishes for circular binaries and grows steeply with eccentricity. For a population of identical binaries the resulting spectrum fluctuation $\\Delta\\Omega^2(f) = \\mathrm{var}\\{h_c^2(f)\\}/\\langle h_c^2(f)\\rangle^2$ scales as $1/N$ and is at most $\\sim10^{-3}$ in the PTA band, too small to detect. When the population is instead sampled from astrophysical merger models with a wide spread in mass and eccentricity, a few bright eccentric binaries dominate and the fluctuations can reach $\\Delta\\Omega^2 \\sim 10^2$, many orders of magnitude above the relative uncertainty $\\sigma_S^2$ of current and planned PTAs. For a single source on a stationary background, the total fluctuation obeys approximately $\\Delta\\Omega^2 \\approx \\rho^4[(G(e)/F(e))^2-1]$, so a binary with $\\rho \\approx 1.4$ and $e=0.5$ gives $\\Delta\\Omega^2$ of order one; the paper argues that such nonstationarity could be the only detectable signature of an otherwise faint eccentric source.","pith_inferences":["Extension: the same variance mechanism should apply to other gravitational wave backgrounds, such as galactic white-dwarf binaries for a space-based detector and extreme-mass-ratio inspirals, where a few bright eccentric sources could produce analogous detectable nonstationarity.","Extension: the paper's single-source approximation $\\Delta\\Omega^2 \\approx \\rho^4[(G(e)/F(e))^2-1]$ implies a sharp dependence on eccentricity; a targeted search for spectral variance in pulsar timing array data could therefore be turned into a measurement of $e$ for the dominant source, which the paper does not develop into a full estimator.","Extension: the neglect of secular frequency evolution over the observation span is the most fragile step; a natural test is to re-run the variance calculation with an evolving orbital frequency over a 10–20 year observation and check whether the harmonic covariance, and hence $\\Delta\\Omega^2$, changes materially.","Extension: if nonstationarity is detected in current or next-generation pulsar timing array data, it would break the degeneracy between a gravitational wave background of circular binaries and an eccentric-binary interpretation, since circular populations produce no such fluctuations."],"forward_implications":["A detection of nonstationarity in the nanohertz gravitational wave background would be evidence that a small number of massive, eccentric supermassive black hole binaries, rather than a smooth uniform population, dominate the signal.","For a homogeneous population of equal binaries the fluctuations are predicted to be undetectable with current pulsar timing arrays, so any observed nonstationarity points to a heterogeneous population or to individual bright sources.","A single eccentric binary with signal-to-noise ratio as low as $\\rho \\approx 1$ can leave a detectable nonstationary imprint even when standard deterministic search methods cannot confidently claim it.","Future pulsar timing arrays with relative uncertainty $\\sigma_S^2 \\approx 8\\times10^{-3}$ should be sensitive to the fluctuations predicted for a massive, highly eccentric population."],"supporting_citations":[{"why":"Supplies the fundamental derivation of GW luminosity from eccentric binaries, including the harmonic decomposition and the eccentricity functions F(e) and g(n,e).","marker":"[1]"},{"why":"Gives the total GW energy radiated by an eccentric binary over its lifetime and the integral used to build the GWB characteristic strain.","marker":"[7]"},{"why":"Introduces the time-fluctuation statistic ΔΩ^2 and the two-frequency correlator used to quantify nonstationarity.","marker":"[8]"},{"why":"Provides the Fisher forecast for PTA relative uncertainty on the GWB amplitude, the benchmark against which fluctuations are judged detectable.","marker":"[9]"},{"why":"Establishes the S(f,t)=g^2(f,t)S0(f) formulation and the relation between measured spectral variance and ΔΩ^2.","marker":"[10]"},{"why":"Provides the theorem linking the GWB characteristic strain to the source population, underpinning the spectrum integrals.","marker":"[11]"},{"why":"Demonstrates that a small subset of bright binaries can dominate the GWB, motivating the realistic-population and single-source calculations.","marker":"[17]"},{"why":"Supplies one of the observation-based SMBHB population models used to sample the realistic populations.","marker":"[21]"},{"why":"Supplies the other population model, providing mass, eccentricity, and redshift distributions for the realistic fluctuation estimates.","marker":"[22]"}],"fun_headline_variants":["Eccentric binaries make gravitational wave background flicker","Faint eccentric binaries leave flicker in GW background","Nonstationary GW background reveals hidden eccentric binaries","Eccentric orbits cause detectable flicker in gravitational waves","GW background flicker from eccentric supermassive binaries"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The calculation replaces the time average of the luminosity over one orbital period with an ensemble average over random orbital phases at the observation time, and it neglects the secular shrinking and speeding-up of the orbit (writing $\\Phi(t)\\approx 2\\pi f_p(t-t_0)$); if binaries evolve appreciably over the observing span, the harmonic correlations on which the variance estimate rests would not be stationary.","fun_headline_variants_meta":{"raw":{"variants":["Eccentric binaries make gravitational wave background flicker","Faint eccentric binaries leave flicker in GW background","Nonstationary GW background reveals hidden eccentric binaries","Eccentric orbits cause detectable flicker in gravitational waves","GW background flicker from eccentric supermassive binaries"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000192,"raw_usage":{"total_tokens":1471,"prompt_tokens":1192,"completion_tokens":279,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":808,"completion_tokens_details":{"reasoning_tokens":203}},"tokens_in":808,"tokens_out":279,"duration_ms":3210,"temperature":1.0,"reasoning_tokens":203,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T00:49:56.634730+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compare the measured variance of the GWB spectrum across independent time segments of a high-cadence pulsar timing array dataset with the variance expected from stationary noise: a source with signal-to-noise ratio $\\rho \\approx 1.4$ and eccentricity $e=0.5$ should produce an excess fluctuation $\\Delta\\Omega^2$ of order one localized at its harmonic frequencies, and a realistic massive eccentric population should produce $\\Delta\\Omega^2$ of order $10^2$; the absence of such excess variance at the predicted level would contradict the paper's central claim.","supporting_citations":[{"cited_title":"The luminosity depends greatly on the eccentricity of the binary through the amplification factor F (e), given by F (e) = 1 + 73 24 e2 + 37 96 e4 (1 − e2)7/2","cited_arxiv_id":null,"evidence_quote":"Supplies the fundamental derivation of GW luminosity from eccentric binaries, including the harmonic decomposition and the eccentricity functions F(e) and g(n,e)."},{"cited_title":"Enoki and M","cited_arxiv_id":null,"evidence_quote":"Gives the total GW energy radiated by an eccentric binary over its lifetime and the integral used to build the GWB characteristic strain."},{"cited_title":"Binary Massive Black Hole Astrophysics","cited_arxiv_id":null,"evidence_quote":"Introduces the time-fluctuation statistic ΔΩ^2 and the two-frequency correlator used to quantify nonstationarity."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the Fisher forecast for PTA relative uncertainty on the GWB amplitude, the benchmark against which fluctuations are judged detectable."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the S(f,t)=g^2(f,t)S0(f) formulation and the relation between measured spectral variance and ΔΩ^2."},{"cited_title":"G(e) F (e) 2 − 1 # = X n,m g(n, e)g(m, e)(⟨fn(t, e)fm(t, e)⟩ −1) ⇔","cited_arxiv_id":null,"evidence_quote":"Provides the theorem linking the GWB characteristic strain to the source population, underpinning the spectrum integrals."},{"cited_title":"Antoniadis et al., Astronomy & Astrophysics 685, A94 (2024)","cited_arxiv_id":null,"evidence_quote":"Demonstrates that a small subset of bright binaries can dominate the GWB, motivating the realistic-population and single-source calculations."},{"cited_title":"Bonetti, A","cited_arxiv_id":null,"evidence_quote":"Supplies the other population model, providing mass, eccentricity, and redshift distributions for the realistic fluctuation estimates."}],"review_version":1}