{"id":"6c2e1aa9-5055-4e80-89e8-8cfcdddff8ca","arxiv_id":"2411.15506","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Using the null stream of a triangular Einstein Telescope, overlapping glitches can be reconstructed and subtracted without signal contamination, preserving parameter estimation accuracy where a two-L-shaped design shows biases.","lead":"This paper shows how the null stream of Einstein Telescope's triangular design lets researchers isolate and remove noise glitches without touching the gravitational wave signal. That preserves accurate measurements of the source's distance and sky location, a key advantage for third-generation observatories.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The method's central premise is that the ET-Δ null stream exactly cancels the GW signal, but all simulations assume perfectly calibrated, identical detectors; any calibration or response asymmetry leaks signal into the glitch reconstruction and directly biases the cleaned parameter estimates.","rationale":"The reader identified exactly the same weakest assumption: the null stream cancellation requires identical, perfectly aligned, perfectly calibrated detectors, and any asymmetry leaks signal into the glitch reconstruction. I agree this is the single most load-bearing concern. The reason it is load-bearing is that the null stream is not a side diagnostic; it is the data set on which the glitch model is fit, with no signal model present. Hence any leaked signal is misattributed to the glitch and subtracted from ET1, producing a direct bias in the cleaned data and the final parameter posteriors. The paper's demonstrations are all conditional on the ideal geometric cancellation, and no robustness study is provided. The proposed calibration-error injection test would settle whether the concern is quantitative or merely formal: at the paper's moderate per-detector SNR, small calibration errors may or may not matter, and the test determines which. I therefore keep the reader's CONDITIONAL verdict unchanged rather than strengthening or weakening it. Secondary weaknesses, such as the absence of a no-glitch baseline for ET-2L and the unmeasured claim of a tenfold speedup, are real but do not threaten the core null-stream mechanism in the same way.","tokens_in":10876,"tokens_out":8953,"duration_ms":97270,"concrete_test":"Repeat the Δt=0 simulation of Section VI after applying independent calibration response functions C_i(f)=1+δ_i(f) to the three ET-Δ detectors before forming the null stream, with δ_i drawn from a realistic model of 2% amplitude and 2° phase errors (and, as a worst case, 5% amplitude and 5° phase errors). Run the same null-stream glitch reconstruction and Bilby parameter estimation, and compare the resulting credible levels for M_total, D_L, and sky location against the no-glitch benchmark and against the perfectly calibrated result. If the posteriors shift by less than roughly 0.5σ, the exact-cancellation premise is not quantitatively load-bearing at the tested SNR; if they shift by more than roughly 1σ, the central claim must be restated as conditional on calibration accuracy.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The core claim rests on Eq. (2): forming the null stream as (d1+d2+d3)/√3 cancels the GW signal exactly because Σ_i h_i = 0. This is true only for three identical, perfectly calibrated detectors. Real interferometers have frequency-dependent calibration amplitude and phase errors, and their antenna responses are not exactly identical. The consequence is not just a slightly imperfect null stream: the null stream is the sole input to the glitch reconstruction step, and BayesWave is run with no signal model. Any leaked GW signal in the null stream is therefore absorbed into the wavelet model of the 'glitch', subtracted from ET1, and removes a real component of the GW signal. This directly biases the subsequent parameter estimates, which is exactly the failure mode the method is intended to prevent. The paper contains no robustness test of this premise: all simulations use identical ET-D PSDs, perfect calibration, and the same waveform projection, so the favorable results in Figs. 3 and 4 are obtained under the exact condition that makes the null stream signal-free. The claim that the method 'fully prevents contamination' is therefore untested for realistic detector asymmetries, including per-detector PSD differences, calibration errors, and small geometric deviations. This is the most load-bearing assumption because it is prerequisite to the entire mitigation pipeline; the missing ET-2L no-glitch baseline and the unmeasured speedup are secondary concerns.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes a glitch mitigation strategy for the Einstein Telescope's triangular configuration (ET-Δ). The authors exploit the sky-position-independent null stream, which cancels the gravitational-wave signal when the three detector streams are summed, to reconstruct an overlapping glitch with BayesWave using no signal model. The reconstructed glitch is subtracted from the affected detector, and parameter estimation is performed on the cleaned data. Through injections of a high-SNR BBH signal with a simulated blip glitch, the authors report that ET-Δ posteriors closely match a no-glitch benchmark, while an ET-2L configuration with simultaneous signal-plus-glitch modeling shows biased extrinsic parameters. They also claim an order-of-magnitude computational speed-up for the null-stream approach.","tokens_in":11128,"tokens_out":3492,"duration_ms":34860,"significance":"If the method is robust in realistic conditions, it offers a practical and conceptually clean way to handle signal-overlapping glitches in third-generation detectors, directly supporting precision cosmology and fundamental-physics goals for ET. A notable strength is that the demonstration uses standard, reproducible analysis tools (gengli, BayesWave, Bilby, IMRPhenomD) and compares against a no-glitch benchmark, which makes the central mechanism easy to check. The work is not circular: glitch parameters are estimated from the null stream, while the no-glitch benchmark and injected true values serve as external references, and no fitted constants are tuned to produce the headline result. The main risk is that the entire pipeline rests on the exactness of the null-stream cancellation, which is assumed but not tested under realistic calibration and response asymmetries.","major_comments":[{"comment":"The central premise that the null stream contains no gravitational-wave signal depends on the exact equality Σ_i h_i = 0. This holds only for three identical, perfectly calibrated detectors with exactly matched antenna responses and no frequency-dependent calibration errors. In reality, calibration amplitude and phase errors and small geometric asymmetries will leak signal power into the null stream. Because BayesWave is run on the null stream with no signal model, any leaked signal will be absorbed into the reconstructed glitch and subsequently subtracted from ET1, removing a real component of the GW signal and biasing the parameter estimates. The paper contains no robustness test of this premise: all simulations use identical ET-D PSDs, perfect calibration, and the same waveform projection. I request an explicit injection study with realistic per-detector calibration errors (e.g., 1% amplitude and a few degrees of phase error, or frequency-dependent analogs) and slightly unequal detector responses, reporting the resulting bias in D_L and Ω relative to the no-glitch benchmark. Without this test, the claim that the null-stream approach \"fully prevents contamination\" is unsupported.","section":"Section III, Eqs. (2)-(3)"},{"comment":"No no-glitch baseline is shown for the ET-2L configuration. The red ET-2L posteriors are compared only against the ET-Δ no-glitch benchmark, but the comparison that isolates the effect of glitch mismodeling is the ET-2L posterior with the glitch versus the ET-2L posterior without the glitch. A two-interferometer network can have intrinsically weaker or biased sky localization and distance measurements even in the absence of any glitch. If the red posteriors are similar to an ET-2L no-glitch baseline, then the claimed \"significant biases due to mismodeling of glitches\" would be primarily a geometric effect, not evidence against simultaneous modeling. I request adding an ET-2L no-glitch benchmark to Figs. 3 and 4, or otherwise quantifying the glitch-induced degradation relative to the appropriate baseline.","section":"Section VI and Figs. 3-4"},{"comment":"The claim of an \"order of magnitude computational speed-up\" is not supported by any measurement. The paper reports no wall-clock times, no likelihood-evaluation counts, and no convergence diagnostics for either the null-stream BayesWave run or the simultaneous ET-2L run. The dimensionality argument (fewer parameters without a signal model) is plausible, but it does not by itself establish a tenfold speed-up for the full glitch-reconstruction and parameter-estimation pipeline. I request either actual timing measurements with the same hardware and convergence criteria, or a reduction of the claim to a qualitative statement about reduced model dimensionality.","section":"Section VII, Results"},{"comment":"Across all nine Δt instances, the signal and glitch morphologies and the noise realizations remain identical; only the glitch onset time changes. This means the headline result is established for one signal, one glitch morphology, and one noise realization. For a \"third-generation-ready\" claim, I request at least a small ensemble with different noise realizations and a second glitch morphology (e.g., a longer-duration or scattered-light glitch) to verify that the green posteriors consistently align with the no-glitch benchmark and that the ET-2L bias pattern is not particular to this single realization.","section":"Section VI, Simulations"}],"minor_comments":[{"comment":"There are several typos: \"trinagular\" should be \"triangular,\" \"complexiteis\" should be \"complexities,\" \"conincides\" should be \"coincides,\" and \"appraoch\" should be \"approach.\"","section":"Introduction"},{"comment":"The sentence \"This consists of instance where the glitch onset conincides with the merger time...\" is missing an article and should read \"This consists of an instance where...\" or \"These consist of instances where...\".","section":"Section VI, Simulations"},{"comment":"The caption states that blue shows \"the measurements when no glitch is introduced,\" but it would be helpful to state explicitly that this benchmark is for ET-Δ and to clarify whether the blue curves are also used for the ET-2L panels (see major comment 2).","section":"Fig. 3 caption"},{"comment":"The phrase \"For all parameters shown\" could be made more precise by listing the parameters: it appears that only M_total, D_L, and sky localization are shown, while other parameters such as mass ratio and spins are not reported; the discussion of \"quality of parameter measurements\" should state which parameters were checked.","section":"Section VII, Results"},{"comment":"No mention is made of data or code availability. Given that the methods rely on public tools, a statement about reproducibility (e.g., a repository with the injection scripts and configuration files) would strengthen the paper.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope for a gravitational-wave data-analysis journal and the core idea is attractive. The main risk is the untested exactness of the null-stream cancellation under realistic detector asymmetries; if the authors can supply the requested robustness injections and the ET-2L no-glitch baseline, the central claims would be substantially better supported. I would not recommend rejection, because the issue is fixable within the manuscript's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read this paper. It shows something genuinely useful: with ET's triangular geometry, you can form a null stream that cancels the GW signal exactly, run BayesWave on that stream to reconstruct a glitch, subtract it from one detector, and then do parameter estimation. The headline result is that even at Δt=0, the cleaned posteriors match the no-glitch benchmark. That is a real demonstration, and the dimensionality reduction (no simultaneous signal+glitch modeling) gives a plausible speedup. The paper is clearly written and the simulation setup is sensible for an idealized first pass.\n\nBut there are soft spots. The biggest is that the null-stream cancellation is taken as exact. All simulations use identical detectors, identical ET-D PSDs, perfect calibration, and the same waveform projection. Real ET will have frequency-dependent calibration errors and response asymmetries. Any leakage of signal into the null stream gets absorbed into the glitch model and then subtracted from ET1, directly biasing the parameters. The claim that the method 'fully prevents contamination' is too strong without a robustness test. This is fixable: add a run with a few percent calibration error and show the posteriors degrade gracefully or quantify the acceptable tolerance.\n\nThe ET-2L comparison is weakened by the absence of a no-glitch baseline for that configuration. The blue benchmark in Figs. 3 and 4 is only for ET-Δ. So we can't tell whether the ET-2L bias comes from glitch mismodeling or simply from the poorer intrinsic sky localization of two L-shaped detectors. That needs an explicit no-glitch ET-2L run.\n\nThe 'approximately ten times faster' claim is not measured. It's plausible from the reduced dimensionality, but a timing measurement or a stronger theoretical argument would be better. Also, the study uses one glitch morphology, one signal, and one noise realization per Δt — nine points total. That's fine for a proof-of-concept, but not for strong statements about robustness.\n\nThe stress-test note is right that the exact cancellation is load-bearing. But I wouldn't call it a fatal flaw. The paper is a proof-of-principle; the assumption is stated. The authors just shouldn't overclaim. The missing baseline and missing timing are more concrete issues.\n\nOverall, the central mechanism works under the stated assumptions, and the paper is a legitimate contribution to the 3G glitch-mitigation discussion. Send it to peer review. A good referee will ask for the robustness tests and the baseline, but the core result is interesting enough to warrant that revision.","headline":"Clean proof-of-principle that ET's null stream can protect parameter estimation from overlapping glitches, but the paper overclaims robustness and misses a key baseline; worth a serious referee.","tokens_in":11728,"tokens_out":2768,"would_cite":true,"duration_ms":26549,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that Einstein Telescope's null stream can erase overlapping glitches without harming parameter estimation.","keywords":["gravitational waves","Einstein Telescope","null stream","glitch mitigation","BayesWave","parameter estimation","blip glitches","third-generation detectors"],"falsifier":"Simulate an ET-Δ binary-black-hole injection with a blip glitch while imposing a small calibration error on one detector—say a 1% amplitude scale or a 1 ms timing offset—and test whether the null-stream subtraction still yields 90% credible intervals that contain the true luminosity distance and sky location.","tokens_in":10670,"feed_emoji":"🔭","tokens_out":4725,"duration_ms":40653,"temperature":0.7,"pith_summary":"This paper argues that Einstein Telescope's planned triangular configuration (ET-Δ) has a built-in 'null stream': summing the three detector outputs cancels the gravitational-wave signal exactly, leaving only whatever instrumental glitch is present plus noise. The authors feed that null stream to BayesWave, a glitch-characterization tool, reconstruct the glitch in isolation, subtract it from the contaminated detector, and then estimate source parameters from the cleaned data. On simulated binary black-hole signals with blip glitches overlapping at the merger, the recovered distance and sky location match a no-glitch benchmark; the same analysis without a null stream, in the proposed ET-2L design, yields biased extrinsic parameters. The paper also claims this route is roughly ten times cheaper than simultaneous signal-plus-glitch modeling, because the signal model and its parameters no longer enter the glitch reconstruction.","feed_headline":"Null stream lets Einstein Telescope erase overlapping glitches","feed_subtitle":"Glitch removal at the signal peak preserves distance and sky measurements, ten times faster than joint modeling.","key_machinery":"The load-bearing identity is the null stream relation (Eq. 2), $\\vec{d}_{\\mathrm{null}} = (\\vec{d}_1+\\vec{d}_2+\\vec{d}_3)/\\sqrt{3}$, together with the geometric cancellation $\\sum_{i=1}^3 \\vec{h}_i = 0$. It converts the hard two-component inference problem of separating an unknown glitch from a known-but-correlated signal into a one-component inference problem on the glitch alone, eliminating the need for a signal waveform during glitch reconstruction and cutting the model's dimensionality by the eleven or more signal parameters. The glitch is represented by a sum of sine-Gaussian wavelets whose number and parameters are explored with a trans-dimensional Markov chain, and the median reconstruction is then subtracted from the raw detector stream before standard parameter estimation.","core_discovery":"The central claim is that the null stream of ET-Δ, defined by $\\vec{d}_{\\mathrm{null}} = (\\vec{d}_1 + \\vec{d}_2 + \\vec{d}_3)/\\sqrt{3}$, contains no gravitational-wave signal because the geometry of the three triangular detectors makes $\\sum_i \\vec{h}_i = 0$. A blip glitch overlapping a loud binary-black-hole signal can therefore be reconstructed from the null stream alone, subtracted from the affected detector, and the source parameters measured as though no glitch had occurred. In the simulations this holds for glitches whose onset coincides with merger as well as for offsets up to roughly 100 ms, while the alternative ET-2L layout, which has no null stream, produces posterior distributions for luminosity distance and sky location that exclude the true values. The extra benefit is computational: glitch-only inference on a lower-dimensional model runs about ten times faster than joint signal-plus-glitch inference.","pith_inferences":["A natural extension the paper leaves open is the tolerance analysis: quantifying how much calibration mismatch or arm-length inequality between the three ET-Δ detectors can be allowed before signal leakage into the null stream starts biasing the posterior; a dedicated injection campaign with controlled detector offsets would settle this.","The same glitch-isolation trick could be used with any future sky-position-independent null stream, so the results give a concrete design criterion: a triangular detector network is not just more sensitive but materially easier to clean.","Since the null stream also contains no signal, it could serve as a diagnostic channel for detector status and noise characterization in real time, a use the authors do not explicitly develop."],"forward_implications":["If the central claim is right, glitches that land exactly on a merger no longer degrade measurements of luminosity distance and sky location for ET-Δ events.","The order-of-magnitude speed-up means glitch mitigation for third-generation detectors can be applied to the large expected event rate without the cost scaling with the duration and loudness of overlapping signals.","Because the null stream removes the need for a signal model, the method should also protect parameter estimation for poorly modeled sources such as core-collapse supernovae, binary neutron-star post-merger signals, or environments with microlensing or eccentricity.","The comparison implies the ET-Δ triangular design has a decisive practical advantage over the ET-2L two-interferometer design for precision science cases such as dark-siren cosmology and tests of general relativity."],"supporting_citations":[{"why":"Defines the null stream concept on which the whole approach rests.","marker":"[61]"},{"why":"Introduces simultaneous signal-plus-glitch modeling with a known waveform, the method the null stream replaces.","marker":"[25]"},{"why":"Provides the BayesWave glitch characterization methodology (sine-Gaussian wavelet reconstruction) used on the null stream.","marker":"[35]"},{"why":"Supplies the ET design study comparison, including the ET-2L configuration without a null stream.","marker":"[31]"},{"why":"The Einstein Telescope triangular design whose geometry gives the sky-independent null stream.","marker":"[54]"},{"why":"The gengli codebase used to simulate realistic blip glitches.","marker":"[59]"},{"why":"Bilby, the parameter-estimation pipeline used after glitch subtraction.","marker":"[7]"},{"why":"The ET-D sensitivity curve used to generate simulated detector noise.","marker":"[28]"},{"why":"IMRPhenomD waveform used to inject the binary-black-hole signal.","marker":"[38]"}],"fun_headline_variants":["Null stream erases glitches for Einstein Telescope","Einstein Telescope's null stream kills glitches","Null stream fast-tracks glitch removal for ET","ET's null stream scrubs glitches, saves signal","Null stream enables glitch-free parameter estimation"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument rests on the geometric cancellation being exact: the three ET-Δ detectors must be identical in response and orientation so that their summed signals vanish, and any real-world asymmetry would leak gravitational-wave signal into the null stream used for glitch reconstruction.","fun_headline_variants_meta":{"raw":{"variants":["Null stream erases glitches for Einstein Telescope","Einstein Telescope's null stream kills glitches","Null stream fast-tracks glitch removal for ET","ET's null stream scrubs glitches, saves signal","Null stream enables glitch-free parameter estimation"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000515,"raw_usage":{"total_tokens":2532,"prompt_tokens":1009,"completion_tokens":1523,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":625,"completion_tokens_details":{"reasoning_tokens":1464}},"tokens_in":625,"tokens_out":1523,"duration_ms":9597,"temperature":1.0,"reasoning_tokens":1464,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T14:13:29.507873+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate an ET-Δ binary-black-hole injection with a blip glitch while imposing a small calibration error on one detector—say a 1% amplitude scale or a 1 ms timing offset—and test whether the null-stream subtraction still yields 90% credible intervals that contain the true luminosity distance and sky location.","supporting_citations":[],"review_version":1}