Pith. sign in

REVIEW 3 major objections 5 minor 3 cited by

A rotating quadrupole mass placed beneath a torsion bar can inject a high-SNR calibration line into the sub-Hz band, reaching 0.24% systematic accuracy.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-02 21:38 UTC pith:LH5GOUZH

load-bearing objection A credible new idea—torque-coupled GCal for torsion bars—with an unquantified model error that undermines the 0.24% headline until fixed. the 3 major comments →

arxiv 2602.19436 v2 pith:LH5GOUZH submitted 2026-02-23 gr-qc astro-ph.IM

Improving calibration accuracy with torque coupled gravity field calibrator for sub-Hz gravitational wave observation in CHRONOS

Yuki Inoue , Daiki Tanabe , Vivek Kumar This is my paper
classification gr-qc astro-ph.IM PACS 04.80.Nn
keywords gravitational-wave calibrationgravity field calibrator (GCal)torsion-bar detectorCHRONOSsub-Hz bandtorque couplingsystematic uncertaintySNR density
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper tries to establish that a torque-coupled gravity field calibrator (GCal) can solve the longstanding low-signal problem for sub-Hz torsion-bar gravitational-wave detectors. By placing a rotating quadrupole rotor directly beneath the torsion bar, the gravitational field couples to the rotational degree of freedom rather than through a weak displacement-to-rotation chain. For the CHRONOS detector, the predicted strain-equivalent calibration amplitude at 1 Hz is 1.18e-14, with an SNR density of 4.25e3 and a fractional systematic uncertainty of 0.24%. If correct, this would be the first practical, SI-traceable calibration line directly available in the 0.1-10 Hz band.

Core claim

The central claim is that the torque-coupled GCal produces a calibration line whose strain-equivalent amplitude is h_GCal(Ω) ≈ -2τ2/(η_g |F_eff| I Ω²), giving a clean 1/Ω² inertial response well above the torsional resonance. At 1 Hz, with a tungsten rotor, this yields |h_GCal| = 1.18e-14 rad, an SNR density of 4.25e3, and a total fractional systematic error of 0.24%, dominated by geometric alignment uncertainties rather than mass or gravitational-constant uncertainties. This would demonstrate for the first time that a high-SNR, absolute calibration line can be injected directly into the sub-Hz band of a torsion-bar detector.

What carries the argument

The key mechanism is the torque-coupled geometry itself: a rotating quadrupole rotor (two diametrically opposite point masses) is placed directly beneath the torsion bar, so the gravitational interaction drives the rotational degree of freedom directly. The dynamical calculation uses a binomial-series multipole expansion of the inverse-distance potential, leading to a closed-form factorized torque amplitude τ2 = (GmM/2R0) S(ξ), where ξ = 2ab/R0² and S(ξ) encodes higher-order multipole contributions. This analytic form allows transparent parameter derivatives and a first-order error-propagation framework, while the high-frequency approximation (Ω ≫ ω0) reduces the response to the inertia-domi

Load-bearing premise

The central calculation replaces the real extended torsion bar and rotor by two-point-mass models matched only through an effective quadrupole moment, and the paper says higher-order multipole corrections are small without giving a numerical bound; if those corrections exceed roughly 0.1% at the 2 m separation, the claimed 0.24% total error would no longer be justified.

What would settle it

Compute the exact Newtonian torque from the continuous mass distributions of the CHRONOS torsion bar and rotor (e.g., by numerical integration) at the nominal geometry h = 2 m, a = 0.302 m, b = 0.16 m; if the relative difference from the point-mass binomial series exceeds about 2e-3, the central calibration-accuracy claim fails. A simpler test: measure the calibration-line amplitude at two frequencies separated by a decade and check that it scales as 1/Ω²; any significant deviation indicates missing dynamics in the inertial-response model.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

Share X Bluesky LinkedIn Reddit HN

If this is right

  • The calibration line at 2f_rot sits well above the detector noise across the 0.1-10 Hz band, with SNR density exceeding 10³ at 1 Hz for tungsten, allowing sensing-function measurement during normal operation.
  • Because the response is inertia-dominated, the calibration amplitude is independent of torsional spring and damping uncertainties, giving a direct torque-to-strain conversion.
  • The fractional systematic uncertainty is about 0.24% for all three rotor materials, indicating that sub-percent absolute calibration accuracy is realistic and limited mainly by alignment metrology.
  • The calibration-line amplitude and SNR scale linearly with rotor mass density, so rotor material choice provides a practical design lever for other torsion-bar detectors.
  • The analytic torque series with factorized S(ξ) permits systematic optimization of geometry (rotor radius, bar length, separation) to balance signal strength against geometric uncertainties.

Where Pith is reading between the lines

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

  • The paper does not bound the higher-order multipole correction from replacing the extended bar and rotor with two point masses; a full numerical integration of the actual mass distributions at h = 2 m, a = 0.302 m, b = 0.16 m would test whether the 0.24% claim survives.
  • The same torque-coupled geometry could be adapted beyond gravitational-wave detectors, e.g., to calibrate low-frequency rotational sensors or torsion balances, where a known gravitational torque provides an independent reference.
  • A cross-check with two different rotor materials at identical geometry would test the predicted linear mass-density scaling and expose any unmodeled systematics in the torque calculation.
  • The 1/Ω² dependence suggests that the calibration line becomes stronger relative to the inertial response at lower frequencies, so the method may be extendable to even quieter parts of the sub-Hz band if the noise spectrum allows.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper proposes a torque-coupled gravity-field calibrator (GCal) for the sub-Hz torsion-bar detector CHRONOS. The bar and rotor are modeled as two point masses (Eqs. 1-4), the Newtonian potential is expanded in a binomial series, and the torque at twice the rotor frequency is given as the closed series τ₂ = (GmM/R₀)[(3/2)ξ² + ...] (Eq. 26). In the high-frequency limit Ω≫ω₀ the strain-equivalent calibration line is h_GCal ≈ −2τ₂/(η_g |F_eff| I Ω²) (Eq. 38). With CHRONOS parameters the authors report |h_GCal| ≈ 1.18×10⁻¹⁴ at 1 Hz, SNR density 4.25×10³, and a 0.24% fractional systematic error from linear propagation of eight parameter uncertainties (Eq. 51). The paper claims this is the first high-SNR, SI-traceable calibration line injected into the sub-Hz band of a torsion-bar detector.

Significance. If correct, the proposal is a useful advance: it gives a transparent, first-principles Newtonian expression for the calibration torque, a closed-form series that permits analytic derivatives, and an error budget that separates mechanical response from GCal parameters. The torque coefficient was checked and the calculation is not circular—no fitting to data is used. The main risk is that the point-mass model in Eqs. (1)-(4) suppresses the finite-size mass moments of the real bar, and the undefined η_g prevents an unambiguous strain conversion. Both are fixable, but they are central to the claimed 0.24% accuracy and to the numerical calibration amplitude.

major comments (3)
  1. [Section III; Eqs. (1)-(4), (26), (51)] The 0.24% systematic budget omits the model error of replacing the continuous torsion bar (and the rotor) by two point masses. Equation (26) is the exact multipole series for the four-point-mass model, not for the physical bar. The parameter a is fixed only by matching the quadrupole moment, and the next (l=4) mass moment contributes at relative order (a/R₀)² ≈ (0.302/2.03)² ≈ 2.2%, far above 0.24% unless a near-cancellation is shown. The text's assertion that higher multipoles are 'small' is not a bound. Provide an estimate (or an explicit symmetry argument) for the l=4 torque using the actual bar/rotor mass distributions, and include the residual in Eq. (51).
  2. [Eqs. (33), (35), Table II] The strain-equivalent calibration amplitude depends on η_g ≡ I_eff/I, but I_eff is never defined and no numerical value is given anywhere. Since h_GCal = Φ_diff/(η_g |F_eff|) and the Table II amplitudes and SNR densities are derived from it, the numerical values are not reproducible until η_g is specified. Please define I_eff (or state that η_g=1 and justify it), and if it is not exactly 1, propagate its uncertainty in Section IV.
  3. [Section III.A; Eq. for SNR] The SNR density is defined as SNR(Ω)=|h_GCal(Ω)|/S_tot(Ω), with S_tot called the 'total angular noise ASD', while h_GCal is strain-equivalent and Fig. 3 plots strain ASD. This is a unit inconsistency: if S_tot is the angular ASD, the ratio should be |Φ_diff|/S_φ; if S_tot is the strain ASD, the text should say so. Table II also labels 'Amplitude (1 Hz) [rad]' for a quantity introduced as dimensionless strain. Clarify these definitions because the headline SNR values depend on them.
minor comments (5)
  1. [Abstract vs. full text] The abstract gives |h_GCal|=1.16×10⁻¹⁴ and SNR=4.16×10³, while the full text and Table II give 1.18×10⁻¹⁴ and 4.25×10³. Harmonize the numbers.
  2. [Section II.A, Eqs. (18)-(23)] The notation 'C 2n 4n' should be written as the binomial coefficient binom(4n,2n) and defined at first use; otherwise the series is hard to follow.
  3. [Section V.B] 'Equation (II A)' is not an equation reference; it should point to Eq. (19) or Eq. (26).
  4. [Title and Section V.C] There are typos: 'gravit y' in the title and 'GCsl' instead of 'GCal' in Section V.C.
  5. [Section V.A and abstract] The abstract states an improvement of more than an order of magnitude over conventional layouts, but no quantitative comparison with a force-coupled GCal is presented. Add a benchmark calculation or temper the wording.

Circularity Check

0 steps flagged

No significant circularity: h_GCal is derived from first-principles Newtonian torque with no fitted parameters; self-citations are external inputs.

full rationale

The paper's central amplitude h_GCal(Ω) ≈ -2τ2/(η_g|F_eff| I Ω^2) (Eq. 38) follows from Newtonian point-mass potential (Eqs. 1-12), binomial expansion (Eqs. 16-26), torsional oscillator transfer function (Eqs. 28-31), and the definition of strain-equivalent response (Eq. 35). No parameter is fitted to the calibration line; the 0.24% systematic budget (Eq. 51) is analytic propagation of assumed measurement uncertainties. The SNR ratio uses the CHRONOS noise ASD from Refs. [30-32], which is an external sensitivity model rather than an output of this derivation; even if that noise model were revised, h_GCal itself would be unchanged. The self-citations are therefore not load-bearing in a circular sense. The unquantified quadrupole truncation (a/R0)^2 ≈ 2% concern is an accuracy/correctness risk, not circularity: the paper uses an effective quadrupole description as a stated modeling choice, not as an equivalent prediction. No step reduces by construction to its own input.

Axiom & Free-Parameter Ledger

2 free parameters · 6 axioms · 0 invented entities

No new physical entities are introduced. The load-bearing model choices are the quadrupole-equivalent point-mass description (parameter a) and the unstated antenna coupling η_g; both are needed to turn the torque calculation into the quoted strain and SNR numbers.

free parameters (2)
  • Effective torsion-bar half-length a = 0.302 m
    Quadrupole-matched parameter replacing the continuous sapphire bar by two point masses (Eqs. 1–4); value assumed from CHRONOS geometry, not derived in this paper, and absorbs the bar's mass-moment details.
  • Geometrical coupling factor η_g = Not stated (effectively 1 in numerical evaluation)
    Appears in Eq. 35 for h_GCal and in the strain noise conversion, but no numerical value is given anywhere; all quoted h_GCal and SNR numbers assume η_g |F_eff| = 1.
axioms (6)
  • domain assumption Torsion bar and rotor can be modeled as two point masses each (Eqs. 1–4), with the bar's continuous mass distribution represented by an effective half-length a that matches the quadrupole moment.
    Underlies the torque series and the 0.24% systematic claim; neglected higher mass multipoles are asserted small without a numerical bound (Section II.A, Section III).
  • domain assumption Small-offset hierarchy ρ << a,b << R0 permits dropping α, β beyond linear order; the offset then enters only through ρ⊥² in R0.
    Used to keep the torque series in the single parameter ξ; valid for ρ=1 mm vs R0≈2 m.
  • domain assumption Calibration frequencies Ω=2ω ∈ 0.1–10 Hz lie far above the torsional resonance (ω0/2π < 0.01 Hz), so the mechanical susceptibility is χ(Ω) ≈ −1/(IΩ²) (Eq. 36).
    Gives the 1/Ω² response that converts torque directly to strain-equivalent amplitude.
  • domain assumption The CHRONOS total noise spectral density is as reported in refs. [30–32].
    The quoted SNR values (4.25×10³ at 1 Hz) are ratios to this unreproduced sensitivity model.
  • domain assumption The parameter uncertainties in Table I (δh=0.05%, δρ⊥=100%, δm=0.0053%, etc.) represent realistic metrology for the CHRONOS environment.
    Directly sets the 0.24% fractional systematic error.
  • domain assumption Rotor-induced mechanical vibrations are dominated by the fundamental and odd harmonics and thus spectrally separated from the 2f calibration line.
    Used to argue non-gravitational couplings do not contaminate the calibration line; not experimentally verified.

pith-pipeline@v1.3.0-alltime-deepseek · 17447 in / 26856 out tokens · 217463 ms · 2026-08-02T21:38:20.792897+00:00 · methodology

0 comments
Cite this review

Pith. "Pith review of Improving calibration accuracy with torque coupled gravity field calibrator for sub-Hz gravitational wave observation in CHRONOS." pith.science (2026). https://pith.science/paper/LH5GOUZH

@misc{pith2026260219436,
  author       = {Pith},
  title        = {Pith review of: Improving calibration accuracy with torque coupled gravity field calibrator for sub-Hz gravitational wave observation in CHRONOS},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LH5GOUZH}},
  note         = {Machine review of arXiv:2602.19436}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

A fundamental challenge in low-frequency gravitational-wave detectors is the limited signal-to-noise ratio (SNR) of calibration lines, particularly in torsion-bar systems where the response is governed by rotational dynamics. In this work, we resolve this issue by optimizing the geometrical configuration of a torque-coupled gravity field calibrator (GCal), achieving an improvement in calibration-line SNR by more than an order of magnitude compared to conventional layouts. For the Cryogenic sub-Hz cROss torsion-bar detector with quantum NOn-demolition Speed-meter (CHRONOS), the calibration signal appears as a monochromatic line within the $0.1$--$10~\mathrm{Hz}$ band. At $1~\mathrm{Hz}$, the strain-equivalent calibration amplitude reaches $|h_{\rm GCal}| = 1.16 \times 10^{-14}$, corresponding to an SNR density of $|h_{\rm GCal}|/S_h = 4.16 \times 10^{3}$. This demonstrates for the first time that a high-SNR calibration line can be directly injected into the sub-Hz band of a torsion-bar detector. A first-order perturbative error propagation analysis yields a total fractional systematic uncertainty of $\delta h_{\rm GCal}/h_{\rm GCal} = 0.24\%$, dominated by geometric alignment uncertainties, while contributions from mass uncertainties and the gravitational constant remain subdominant. The corresponding absolute systematic uncertainty is $\delta h_{\rm GCal} \sim 10^{-17}$ at $1~\mathrm{Hz}$. These results establish torque-coupled gravitational calibration as a practical solution to the longstanding low-SNR problem in sub-Hz torsion-bar detectors and provide a robust pathway toward precision absolute calibration in the low-frequency regime.

Figures

Figures reproduced from arXiv: 2602.19436 by Daiki Tanabe, Vivek Kumar, Yuki Inoue.

Figure 1
Figure 1. Figure 1: Conceptual comparison between conventional force-c [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Schematic illustration of the torque-coupled GCal configu [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Calculated strain sensitivity of the CHRONOS (red curve) t [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: SNR density of the GCal-induced calibration signal for the C [PITH_FULL_IMAGE:figures/full_fig_p012_4.png] view at source ↗

discussion (0)

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

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Science of Cryogenic sub-Hz cROss torsion bar detector with quantum NOn-demolition Speed meter (CHRONOS)

    astro-ph.IM 2026-04 unverdicted novelty 8.0

    CHRONOS is a proposed cryogenic torsion-bar detector with quantum non-demolition speed-meter readout targeting 10^{-18} strain sensitivity at 2 Hz to open the sub-Hz gravitational-wave window from the ground.

  2. Prospects for Observing Gravity-gradient Noise and Earthquake Gravity Signals with CHRONOS

    physics.ins-det 2026-06 unverdicted novelty 5.0

    CHRONOS is projected to detect prompt gravitational signals from Mw 5.2 earthquakes within ~90 km (SNR ~3.62 at 40 km in sub-Hz band) while Rayleigh-wave Newtonian noise dominates below ~0.5 Hz.

  3. Probing Yukawa Gravity with Modulated Newtonian Cancellation in the CHRONOS Detector

    gr-qc 2026-04 unverdicted novelty 5.0

    Torsion-bar detector with differential mass cancellation reaches |α_Y| = 2.4×10^{-5} at λ = 8 m for Yukawa gravity deviations, limited by source-mass geometry uncertainties after ~26 hours.

Reference graph

Works this paper leans on

41 extracted references · 7 linked inside Pith · cited by 3 Pith papers

  1. [1]

    B. P. Abbott et al. , Physical Review Letters 116, 061102 (2016)

    2016

  2. [2]

    Aasi et al

    J. Aasi et al. (LIGO Scientific Collaboration), Classical and Quantum Gravity 32, 074001 (2015) , arXiv:1411.4547 [gr-qc]

    Pith/arXiv arXiv 2015

  3. [3]

    Acernese et al

    F. Acernese et al. , Class. Quantum Grav. 32, 024001 (2015)

    2015

  4. [4]

    Aso et al

    Y. Aso et al. , Phys. Rev. D 88, 043007 (2013)

    2013

  5. [5]

    B. P. Abbott et al. , Physical Review X 9, 031040 (2019)

    2019

  6. [6]

    Abbott et al

    R. Abbott et al. , Physical Review X 11, 021053 (2021)

    2021

  7. [7]

    Abbott et al

    R. Abbott et al. , Physical Review X 13, 041039 (2023)

    2023

  8. [8]

    Inoue et al

    Y. Inoue et al. , (2018), arXiv:1804.08249 [gr-qc]

    Pith/arXiv arXiv 2018

  9. [9]

    Campeti, E

    P. Campeti, E. Komatsu, D. Poletti, and C. Baccigalupi, Journal of Cosmology and Astroparticle Physics , 012 (2021) , arXiv:arXiv:2007.04241 [astro-ph.CO]

    Pith/arXiv arXiv 2021

  10. [10]

    S. M. Feeney et al. , (2018), arXiv:1802.03404

    Pith/arXiv arXiv 2018

  11. [11]

    A. G. Riess et al. , Astrophys. J. 826, 56 (2016)

    2016

  12. [12]

    Nissanke et al

    S. Nissanke et al. , Astrophys. J. 725, 496 (2010)

    2010

  13. [13]

    B. P. Abbott et al. , Nature 551, 85 (2017)

    2017

  14. [14]

    Kuck, Responsivity of detectors for radiant power of lasers, Fina l Report, Tech

    S. Kuck, Responsivity of detectors for radiant power of lasers, Fina l Report, Tech. Rep. EUROMET.PR–S2 (EUROMET Comparison Project 156, 2009)

    2009

  15. [15]

    B. P. Abbott et al. , Physical Review D 95, 062003 (2017)

    2017

  16. [16]

    Tuyenbayev et al

    D. Tuyenbayev et al. , Classical and Quantum Gravity 34, 015002 (2016)

    2016

  17. [17]

    Cahillane et al

    C. Cahillane et al. , Phys. Rev. D 96, 102001 (2017)

    2017

  18. [18]

    E. D. Hall et al. , Classical and Quantum Gravity 35, 065003 (2018)

    2018

  19. [19]

    Inoue et al

    Y. Inoue et al. , Review of Scientific Instruments (2023), arXiv:2302.12180 [gr-qc]

    arXiv 2023

  20. [20]

    Goetz et al

    E. Goetz et al. , Class. Quantum Grav. 26, 245011 (2009)

    2009

  21. [21]

    Karki et al

    S. Karki et al. , Review of Scientific Instruments 87, 114503 (2016)

    2016

  22. [22]

    Mossavi et al

    K. Mossavi et al. , Phys. Lett. A 353, 1 (2006)

    2006

  23. [23]

    KAGRA Collaboration, Progress of Theoretical and Expe rimental Physics , 05A102 (2021), arXiv:2009.09305 [gr-qc]

    arXiv 2021

  24. [24]

    Matone et al

    L. Matone et al. , Class. Quantum Grav. 24, 2217 (2007)

    2007

  25. [25]

    Raffai et al

    P. Raffai et al. , Phys. Rev. D 84, 082002 (2011)

    2011

  26. [26]

    K. Oide, K. Tsubono, and H. Hirakawa, Jpn. J. Appl. Phys. 19, L123 (1980)

    1980

  27. [27]

    Est´ evezet al

    D. Est´ evezet al. , Classical and Quantum Gravity 35, 235009 (2018)

    2018

  28. [28]

    Acernese et al

    F. Acernese et al. , Classical and Quantum Gravity 35, 205004 (2018)

    2018

  29. [29]

    Akutsu et al

    T. Akutsu et al. , Progress of Theoretical and Experimental Physics 2021, 05A101 (2021)

    2021

  30. [30]

    Inoue et al

    Y. Inoue et al. , (2025), 10.48550/arXiv.2509.23172, arXiv:2509.23172 [astro-ph.IM]

    doi:10.48550/arxiv.2509.23172 2025

  31. [31]

    Inoue, D

    Y. Inoue, D. Tanabe, M. A. Ismail, V. Kumar, M. J. S. Ongla o, and T.-C. Yu, (2025), 10.48550/arXiv.2510.24780 , arXiv:2510.24780 [physics.ins-det]

    Pith/arXiv arXiv doi:10.48550/arxiv.2510.24780 2025

  32. [32]

    Torque cancellation effect of intensity noise for a cryoge nic torsion-bar interferometer,

    D. Tanabe et al. , “Torque cancellation effect of intensity noise for a cryoge nic torsion-bar interferometer,” (2025), arXiv:2510.24779 [gr-qc] . 19

    arXiv 2025

  33. [33]

    Ando et al

    M. Ando et al. , Physical Review Letters 105, 161101 (2010)

    2010

  34. [34]

    L. Ju, C. Zhao, and D. Blair, in Proceedings of the 13th Marcel Grossmann Meeting (2019) pp. 2487–2492

    2019

  35. [35]

    R. X. Adhikari, G. Gonzalez, M. Landry, and B. O’Reilly ( LIGO Scientific Collaboration), Classical and Quantum Gravity 20, S903 (2003)

    2003

  36. [36]

    M. Wang, C. Bond, F. Br¨ uckner, et al. , Physical Review D 87, 096008 (2013) , arXiv:1303.5236 [gr-qc]

    Pith/arXiv arXiv 2013

  37. [37]

    Chen, Physical Review D 67, 122004 (2003) , arXiv:gr-qc/0208051

    Y. Chen, Physical Review D 67, 122004 (2003) , arXiv:gr-qc/0208051

    Pith/arXiv arXiv 2003

  38. [38]

    Harms, Living Reviews in Relativity 18, 3 (2015)

    J. Harms, Living Reviews in Relativity 18, 3 (2015)

    2015

  39. [39]

    P. R. Saulson, Fundamentals of Interferometric Gravitational Wave Detec tors (World Scientific, 1994)

    1994

  40. [40]

    Cg-6000,

    “Cg-6000,” http://www.vibra.co.jp

  41. [41]

    Inoue et al

    Y. Inoue et al. , Appl. Opt. 55, D22 (2016)

    2016