Pith. sign in

REVIEW 2 major objections 4 minor 64 references

Low multipole mapmaking for global 21-cm experiments

T0 review · 2 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A CMB-style mapmaking inversion can recover the 21-cm global signal from chromatic-beam data that break standard foreground fits.

desk verdict Solid methods paper: CMB-style GLS mapmaking for global 21-cm works when the high-ℓ foreground mean/covariance are right, but the robustness claim is tested only against scatter, not bias. read the letter →

arxiv 2506.21258 v1 pith:TRV5HNNV submitted 2025-06-26 astro-ph.CO

Yordan D. Ignatov , Jonathan R. Pritchard This is my paper · ORCID
classification astro-ph.CO
keywords 21-cmcosmologyglobalsignalbeamchromaticityforegroundsubtractionmapmakingsphericalharmonicdecompositionepochofreionizationcosmicdawn
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper argues that the standard failure mode of global 21-cm experiments, spectral oscillations produced when a frequency-dependent antenna beam averages over a spatially structured foreground, can be handled before foreground fitting by treating the multi-antenna observation as a linear inverse problem. Writing the beam-weighted sky as $d = A a + n$, the method inverts for the low-order spherical-harmonic coefficients using generalized least squares, with the unmeasured high-$\ell$ foreground modes accounted for through their assumed mean and covariance. In simulations with chromatic beams up to curvature $c = 3.4 \times 10^{-2}$ and a foreground prior with 10\% uncertainty, the recovered monopole yields unbiased 21-cm posteriors, while a beam-factor chromaticity-corrected single-spectrum fit either loses the signal or returns extremely broad contours. The payoff, if the method works on real data, is a signal-extraction pipeline that requires only a probabilistic foreground prior rather than an exact foreground template.

What carries the argument

The load-bearing object is the linear observation equation $d = G P Y B a + n$, where $B$ encodes the azimuthally symmetric beam as spherical-harmonic coefficients, $Y$ maps coefficients to pixels, $P$ is the pointing matrix of a latitudinally separated antenna array, and $G$ bins the time-ordered data. The method inverts this equation with the generalized least-squares estimator $\hat{a}' = W'(d - A''\mu'')$, where the covariance matrix $C = S + N$ includes both noise and the uncertainty in the higher-order modes. The mean $\mu''$ and covariance $C''_a$ of the corrected modes come from a stochastic 2-basemap extrapolation (S2BE) model: Gaussian deviations are added to the spectral index of every pixel, so the fractional pixel error is roughly Gaussian at the percent level, and the covariance propagates analytically through the spherical-harmonic transform.

What would settle it

Generate mock observations from a foreground sky whose high-$\ell$ modes have the same mean and covariance as the S2BE prior but are drawn from a non-Gaussian or spatially correlated distribution, or whose mean is shifted by a 10\% calibration error in the 408 MHz base map, then run the mapmaking pipeline with the unmodified S2BE prior and check whether the recovered 21-cm amplitude and width remain within $1\sigma$ of the injected values.

Watch

Extended reading notes

Core claim

The central claim is that the beam-chromatic foreground coupling that biases single-spectrum fits is removable in advance: a global 21-cm experiment can estimate the chromaticity-free low-$\ell$ sky modes, including the monopole, by inverting the observation equation with generalized least squares. The inversion is under-determined for high-$\ell$ modes because the beams are broad and the antennas fixed, so the method subtracts $A''\mu''$ from the data and inflates the noise covariance by $S = A''C''_a A''^T$, using only the mean and covariance of modes above $\ell_{\mathrm{mod}} = 5$. With these two inputs, the recovered monopole is fit by a smooth log-polynomial foreground and a Gaussian 21-cm model. The paper demonstrates that for chromatic beams with curvature $c = 1.6 \times 10^{-2}$ and $3.4 \times 10^{-2}$, the method recovers the injected 21-cm signal to $1\sigma$ with 10\% foreground-correction error, where the BFCC single-spectrum fit fails; even at $c = 5.2 \times 10^{-2}$ the signal is recovered, while 20\% correction error widens the posterior and weakens the detection.

Load-bearing premise

The load-bearing premise is that the mean and covariance of the high-order foreground modes ($\ell > 5$) are known and that the true high-$\ell$ foregrounds are a Gaussian draw from that distribution; in the paper's simulations the correction is built from the exact S2BE model that generated the data, so a systematically biased or structurally different prior is never tested.

Editorial extensions

If this is right

  • For chromatic beams up to $c = 3.4 \times 10^{-2}$, the mapmaking monopole yields unbiased 21-cm posteriors at 10\% foreground-correction error, while the BFCC single-spectrum fit yields posteriors consistent with no signal or spanning roughly 50 to 850 mK.
  • With a 7-antenna latitudinally separated array, low-order modes up to $\ell_{\mathrm{mod}} = 5$ can be reconstructed with monopole error near the 2 mK noise floor, and the recovered monopole can be fit with a 3-term log-polynomial.
  • The method requires only the mean and covariance of $\ell > 5$ foreground modes, not an exact foreground template, so a calibrated probabilistic foreground model with error propagation could supply these for real data.
  • At 20\% foreground-correction error the recovered 21-cm signal is still inferred at $2\sigma$, but the posterior is broad enough that it would be difficult to distinguish from unknown systematics.
  • The same observation-equation formalism extends to non-azimuthally symmetric beams, antennas with different beam profiles, and pointable or satellite-based instruments without changing the mapmaking inversion.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the approach transfers to real instruments, the natural first validation is on data from existing latitudinally separated sites, where beam calibration errors, assumed perfect here, will be the dominant new uncertainty.
  • The 10\% tolerance probably depends on the Gaussianity of the high-$\ell$ modes; a prior with the same mean and covariance but heavier tails or spatial correlations could bias the monopole estimate more than the quoted 10\%.
  • A satellite-based or steerable full-sky scanning experiment would remove the need for the high-$\ell$ prior entirely, since every mode would be observed; the same observation-equation formalism would apply unchanged.
  • The method could be combined with training-set foreground models to absorb beam uncertainty, since the mapmaking inversion currently assumes the beam spherical harmonics are known exactly.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 4 minor

Summary. The paper presents a mapmaking-based method for extracting the sky-averaged 21-cm signal from global 21-cm experiments. The authors write the beam-weighted observation of the sky by multiple latitudinally separated antennas as a linear observation equation, invert it with generalized least squares to estimate low-order spherical harmonics (ℓ ≲ 5), and correct for higher-order modes using a known mean and covariance of a stochastic foreground model. They benchmark the method against a single-spectrum fit with beam-factor chromaticity correction (BFCC), showing that BFCC fails for chromatic beams while the mapmaking method recovers the injected 21-cm signal when the higher-mode foreground correction has 10% scatter. The method is demonstrated on controlled simulations in which the sky is a realization of the same S2BE model whose mean and covariance provide the correction.

Significance. If the result holds, the paper offers a conceptually new route to mitigating beam-chromatic foreground coupling in global 21-cm experiments: instead of requiring an accurate foreground template, it uses a probabilistic description of high-ℓ foreground modes. The GLS derivation is standard and correct, the S2BE mean and covariance are derived analytically in Appendix A, and the code is made publicly available. The comparison with BFCC is useful and fairly constructed. The main limitation, correctly identified by the authors in the abstract and Section 6.2, is that the method assumes the mean and covariance of high-ℓ foreground modes are known; the demonstrations are self-consistent in the sense that the same model generates the sky and provides the correction. This makes the headline '10% uncertainty' claim conditional on the correction being correctly centered as well as correctly covarianced, a point that deserves explicit stress testing before the method can be taken as robust in realistic settings.

major comments (2)
  1. [Sections 5.3-5.4, Eqs. (27)-(29)] The 10%-error demonstrations are matched-filter tests: the sky is generated from the S2BE model and the correction uses the exact analytic mean μ″ and covariance C″_a of that same model. A biased correction mean, for example from base-map calibration errors of order 10-60% as cited in Section 5, would enter Eq. (26) as the deterministic leakage term W′A″(μ″_true − μ″_wrong), which is not controlled by the covariance S. Since Section 5.1 shows that uncorrected high-ℓ modes can bias the monopole at the ~K level, this leakage could easily exceed the ~20 mK posterior width reported in Fig. 8. The paper should include simulations with a systematically biased μ″ (e.g., a spectral-index offset or an independent foreground model) and report the resulting monopole bias and posterior coverage; without such a test, the statement that the method succeeds with 'uncertainty at the 10% level' is only supported for correctly centered scatter.
  2. [Section 6.2] The paper acknowledges that non-Gaussian high-ℓ modes would break the optimality of the GLS inversion, but no non-Gaussian or structurally different foreground model is tested. The S2BE model produces approximately Gaussian Δ″ by construction, as shown in Eq. (31), so the simulations never exercise the regime where the assumed Gaussian likelihood is misspecified. A concrete test with a non-Gaussian foreground residual (or a foreground model with spatial correlations not captured by the diagonal pixel covariance) would quantify how much degradation actually occurs and whether the method remains usable at the claimed 10% error level.
minor comments (4)
  1. [Section 3.2.3] The sentence 'the length scale of each pixel must be greater than the size of the smallest angular scale represented by lmax' should read 'smaller than'; the subsequent justification (NSIDE=32 gives a pixel width about 3 times smaller than the lmax=32 scale) is consistent with the corrected wording.
  2. [Section 3.2.3] The phrase 'approximately an order of 5 smaller' appears to be a typo; from the context and Fig. 2, the residual at lmax=32 is about an order of magnitude smaller than the 21-cm signal, not five orders of magnitude.
  3. [Fig. 10 caption] The caption refers to 'c = 3.2 × 10−2' but the corresponding cases in the text and Fig. 8 use c = 3.4 × 10−2; the values should be made consistent.
  4. [Fig. 10 caption] The phrase 'the three right-hand cases in Figs. 7 to 9' is unclear; it should specify the exact panels (e.g., Fig. 7 right, Fig. 8 right, Fig. 9 right).

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the 21-cm signal is injected independently and never used to build the high-ell foreground correction; the GLS covariance is an assumed input, not a fitted parameter reused as a prediction.

full rationale

The derivation chain is not circular. The S2BE foreground model is used both to generate the mock sky and to construct the mean mu'' and covariance C''_a of the high-ell modes (Eqs. 28-29, Appendix A); the GLS estimator (Eq. 26) then uses these as assumed inputs, exactly as a CMB mapmaker uses a noise covariance. The 21-cm signal is added separately (Eq. 14, Section 2.1) and is never used to set mu'', C''_a, or S; it is only recovered by the final single-spectrum fit. No parameter is fitted to the data and then re-used as a prediction: the recovered monopole is fit with a log-polynomial plus the 21-cm model, and the 21-cm parameters are the output. The only caveat is that the simulations are self-consistent: the correction mean and covariance coincide with the generative model, and non-Gaussian or biased high-ell modes are not exercised (Section 6.2). This is a validation limitation and a robustness risk, but it is not circularity: the mapmaking success is not equivalent by construction to the assumptions, because the 21-cm inference step is a non-trivial fit to noisy data and could in principle fail even with a correct foreground covariance. Citations to the authors' prior work (e.g., Ignatov et al. 2024 for dipole information) are not load-bearing here; the need for multiple antennas is also demonstrated directly in Fig. 5. No specific circular step can be quoted.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The method's central dependence is on the assumed Gaussian mean and covariance of high-l foreground modes (Section 5.2). The free parameters (sigma_delta, l_mod, N_ant) are chosen by hand for the demonstration; the S2BE model itself is a new but simple model that simulates pixel noise without calibration errors (Section 5.3).

free parameters (3)
  • sigma_delta (equiv. sigma_T at 70 MHz) = 0.057 (10%), 0.11 (20%)
    Sets the scatter of the S2BE foreground realization about its mean; the central positive result (10%) and failure (20%) depend on this choice (Section 5.3).
  • l_mod = 5
    Truncation order for estimated low-order modes, chosen conservatively from Fig. 5 for N_ant=7 to avoid monopole bias from poorly constrained modes.
  • N_ant = 7
    Number of antennas in the simulated global array; the method fails with fewer antennas (Fig. 5), so the demonstration requires this array configuration.
assumptions (5)
  • domain assumption The antenna beam is azimuthally symmetric.
    Used to write the beam convolution as a multiplication by a diagonal matrix in harmonic space (Eq. 6).
  • ad hoc to paper The high-l (l > l_mod) foreground modes are Gaussian with known mean and covariance.
    Central to the GLS correction (Eqs. 24-26); only approximately true for the S2BE model (Eq. 31) and untested for real skies.
  • domain assumption The 21-cm signal is a pure monopole.
    Higher-order multipoles of the signal are neglected (Eq. 14).
  • domain assumption Noise is Gaussian, independent, with radiometer equation variance from Eq. (11).
    Standard assumption for total-power radiometers; enables the closed-form GLS solution.
  • ad hoc to paper The true foreground sky is a realization of the S2BE model used for correction.
    The simulation and correction share the same generative model (Section 5.3-5.4); not a property of the real sky, which may have a different foreground structure.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Low multipole mapmaking for global 21-cm experiments." pith.science (2026). https://pith.science/paper/TRV5HNNV

@misc{pith2026250621258,
  author       = {Pith},
  title        = {Pith review of: Low multipole mapmaking for global 21-cm experiments},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TRV5HNNV}},
  note         = {Machine review of arXiv:2506.21258}
}
abstract

The 21-cm global signal is obscured by very bright galactic and extra galactic foreground emissions. Typical single-spectrum fit (SSF) based methods for foreground/signal separation can result in biased estimates of the cosmological signal due to the presence of spectral oscillations induced by the interaction between chromatic beams and the spatial shape of the foregrounds. Modelling this interaction requires some amount of assumed foreground information. We present a mapmaking-based approach which describes the beam-weighted observation of the sky by multiple globally-distributed antenna experiments as an observation equation. This equation is inverted in order to estimate the low-order sky modes ($\ell\lesssim10$). The resulting chromaticity-free sky monopole is then fit with a smooth foreground function and a 21-cm model. Given the insensitivity of global 21-cm experiments to small angular scales, we rely on the mean and covariance of higher-order foreground modes being known. We show that this mapmaking-based method is capable of inferring the cosmological signal in cases where a SSF with a simple beam-factor based chromaticity correction fails, even when the foreground model used in the mapmaking method features uncertainty at the 10\% level.

Figures

Figures reproduced from arXiv: 2506.21258 by the authors.

Figure 1
Figure 1. 3.2.3 Pixel resolution and 𝑁𝑙max In order to use equation (10) to generate mock observations of the 21-cm sky, the foreground and 21-cm skies must be represented as a vector of spherical harmonic coefficients a (𝑖) for each frequency, equal to the sum of the foreground spherical harmonics a (𝑖) fg and 21- cm monopole spherical harmonics a (𝑖) 21 . These spherical harmonic vectors are given by a (𝑖) fg = 𝑌 −1T (𝑖) fg… view at source ↗
Figure 2
Figure 2. The RMS temperature difference between observations of the 2BE foreground sky truncated at 𝑙max and 2𝑙max. Noiseless mock observations at 70 MHz are used, generated with the standard observing strategy and binned into a single time bin. This RMS difference is compared to the approximate 21-cm signal scale (dotted line). beam shape at the reference frequency 𝑓beam(n; 𝜈ref), i.e. as if the experiment were achromatic. … view at source ↗
Figure 3
Figure 3. Results of the BFCC-corrected single-spectrum fit for data form a 7- antenna global array, collected with achromatic beams and fit with 𝑁poly = 3 (top-left panel), a chromaticity curvature of 𝑐 = 1.6 × 10−2 and 𝑁poly = 5 (top-right panel) and chromaticity curvature of 𝑐 = 3.4 × 10−2 and 𝑁poly = 6 (bottom-left panel). For each panel, the upper sub-figure shows the coloured 1𝜎 and 2𝜎 iso-probability contours of the fu… view at source ↗
Figures from the paper (6 more)
Figure 5
Figure 5. Figure 5: Left: the absolute reconstruction error of the foreground monopole at 70 MHz, when the observed foregrounds are truncated at 𝑙mod and a max￾imum likelihood reconstruction of order 𝑙mod is used. Shown are the results for estimation based on data taken from 3 (dash-dot),…
Figure 4
Figure 4. Figure 4: Mollweide projection of the beam-convolved foreground sky at 60 MHz in Galactic coordinates for 𝑙 ≤ 5 (top) and 5 < 𝑙 ≤ 32 (bottom). Overlaid are the drift-scan sky tracks of 7 antennas, equally spaced in latitude every 26◦ (red dashed lines). error is given by |𝑎ˆ (70…
Figure 6
Figure 6. Figure 6: Left: the fractional standard deviation of the S2BE model in the Gaussian approximation. The lines correspond to a percentage error of 10% and 20% (bottom to top) at 70 MHz. Right: percentage deviation from the mean of each pixel of a realisation of the S2BE model at 7…
Figure 7
Figure 7. Figure 7: Results from the mapmaking method for data collected with an achromatic beam, where the contribution from modes above 𝑙mod = 5 have been corrected for perfectly with 𝜎𝑇 (70 MHz) = 0% (left) and imperfectly with 𝜎𝑇 (70 MHz) = 10% (right). For both scenarios, an 𝑁poly = …
Figure 9
Figure 9. Figure 9: Results from the mapmaking method (c.f [PITH_FULL_IMAGE:figures/full_fig_p012_9.png]
Figure 11
Figure 11. Figure 11: A flowchart showing the mapmaking 21-cm signal extraction method. Absolute-calibrated timeseries spectral data from a global array of 21-cm experiments is ran through a multipole estimation step, which recovers the monopole of the 21-cm signal and foregrounds. This st…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

64 extracted references · 18 canonical work pages

  1. [1]

    Anstey D., de Lera Acedo E., Handley W., 2021, @doi [ ] 10.1093/mnras/stab1765 , https://ui.adsabs.harvard.edu/abs/2021MNRAS.506.2041A 506, 2041

    work page doi:10.1093/mnras/stab1765 2021

  2. [2]

    Anstey D., de Lera Acedo E., Handley W., 2023, @doi [ ] 10.1093/mnras/stad156 , https://ui.adsabs.harvard.edu/abs/2023MNRAS.520..850A 520, 850

    doi:10.1093/mnras/stad156 2023

  3. [3]

    Barkana R., 2018, @doi [ ] 10.1038/nature25791 , 555, 71

    doi:10.1038/nature25791 2018

  4. [4]

    Barrett J., Cappallo R., Soohoo J., 2025, EDGES Memo 396, MIT Haystack Observatory, https://dspace.mit.edu/bitstream/handle/1721.1/154786/memo_EDGES_396.pdf

    work page 2025

  5. [5]

    J., 2015, @doi [ ] 10.1088/0004-637x/799/1/90 , 799, 90

    Bernardi G., McQuinn M., Greenhill L. J., 2015, @doi [ ] 10.1088/0004-637x/799/1/90 , 799, 90

    doi:10.1088/0004-637x/799/1/90 2015

  6. [6]

    D., Rogers A

    Bowman J. D., Rogers A. E. E., Monsalve R. A., Mozdzen T. J., Mahesh N., 2018, @doi [ ] 10.1038/nature25792 , https://ui.adsabs.harvard.edu/abs/2018Natur.555...67B 555, 67

    doi:10.1038/nature25792 2018

  7. [7]

    Brandenberger R., Cyr B., Shi R., 2019, @doi [JCAP] 10.1088/1475-7516/2019/09/009 , https://ui.adsabs.harvard.edu/abs/2019JCAP...09..009B 2019, 009

    doi:10.1088/1475-7516/2019/09/009 2019

  8. [8]

    arXiv:2410.00076

    Bull P., et al., 2024, @doi [arXiv e-prints] 10.48550/arXiv.2410.00076 , https://ui.adsabs.harvard.edu/abs/2024arXiv241000076B p. arXiv:2410.00076

    work page doi:10.48550/arxiv.2410.00076 2024

Show all 64 references
  1. [9]

    O., 2021, @doi [Philosophical Transactions of the Royal Society of London Series A] 10.1098/rsta.2019.0564 , https://ui.adsabs.harvard.edu/abs/2021RSPTA.37990564B 379, 20190564

    Burns J. O., 2021, @doi [Philosophical Transactions of the Royal Society of London Series A] 10.1098/rsta.2019.0564 , https://ui.adsabs.harvard.edu/abs/2021RSPTA.37990564B 379, 20190564

    2021

  2. [10]

    C., Barrett J

    Cappallo R. C., Barrett J. P., 2025, EDGES Memo 470, MIT Haystack Observatory, https://www.haystack.mit.edu/wp-content/uploads/2025/01/EDGES-MEMO_470.pdf

    2025

  3. [11]

    Carter G., Handley W., Ashdown M., Razavi-Ghods N., 2025, @doi [arXiv e-prints] 10.48550/arXiv.2501.01417 , https://ui.adsabs.harvard.edu/abs/2025arXiv250101417C

    2025 doi

  4. [12]

    Chen X., Yan J., Deng L., Wu F., Wu L., Xu Y., Zhou L., 2021, @doi [Philosophical Transactions of the Royal Society of London Series A] 10.1098/rsta.2019.0566 , https://ui.adsabs.harvard.edu/abs/2021RSPTA.37990566C 379, 20190566

    2021

  5. [13]

    Cohen A., Fialkov A., Barkana R., Lotem M., 2017, @doi [ ] 10.1093/mnras/stx2065 , https://ui.adsabs.harvard.edu/abs/2017MNRAS.472.1915C 472, 1915

    2017 doi

  6. [14]

    A., 2018, @doi [ ] 10.3847/2041-8213/aae318 , https://ui.adsabs.harvard.edu/abs/2018ApJ...866L...7D 866, L7

    Deshpande A. A., 2018, @doi [ ] 10.3847/2041-8213/aae318 , https://ui.adsabs.harvard.edu/abs/2018ApJ...866L...7D 866, L7

    2018 doi

  7. [15]

    B., Schinzel F

    Dowell J., Taylor G. B., Schinzel F. K., Kassim N. E., Stovall K., 2017, @doi [ ] 10.1093/mnras/stx1136 , https://ui.adsabs.harvard.edu/abs/2017MNRAS.469.4537D 469, 4537

    2017 doi

  8. [16]

    B., 1958, @doi [Proc

    Field G. B., 1958, @doi [Proc. IRE] 10.1109/JRPROC.1958.286741 , https://ui.adsabs.harvard.edu/abs/1958PIRE...46..240F 46, 240

    1958

  9. [17]

    W., Lang D., Goodman J., 2013, @doi [ ] 10.1086/670067 , https://ui.adsabs.harvard.edu/abs/2013PASP..125..306F 125, 306

    Foreman-Mackey D., Hogg D. W., Lang D., Goodman J., 2013, @doi [ ] 10.1086/670067 , https://ui.adsabs.harvard.edu/abs/2013PASP..125..306F 125, 306

    2013 doi

  10. [18]

    Fraser S., et al., 2018, @doi [Phys. Lett. B] 10.1016/j.physletb.2018.08.035 , https://ui.adsabs.harvard.edu/abs/2018PhLB..785..159F 785, 159

    2018 doi

  11. [19]

    Haslam C. G. T., Salter C. J., Stoffel H., Wilson W. E., 1982, , https://ui.adsabs.harvard.edu/abs/1982A&AS...47....1H 47, 1

    1982

  12. [20]

    D., Puchwein E., 2018, @doi [ ] 10.1038/s41586-018-0796-5 , https://ui.adsabs.harvard.edu/abs/2018Natur.564E..32H 564, E32

    Hills R., Kulkarni G., Meerburg P. D., Puchwein E., 2018, @doi [ ] 10.1038/s41586-018-0796-5 , https://ui.adsabs.harvard.edu/abs/2018Natur.564E..32H 564, E32

    2018 doi

  13. [21]

    C., Ahn K., 2024, @doi [ ] 10.3847/1538-4357/ad2209 , https://ui.adsabs.harvard.edu/abs/2024ApJ...964...21H 964, 21

    Hotinli S. C., Ahn K., 2024, @doi [ ] 10.3847/1538-4357/ad2209 , https://ui.adsabs.harvard.edu/abs/2024ApJ...964...21H 964, 21

    2024 doi

  14. [22]

    D., Pritchard J

    Ignatov Y. D., Pritchard J. R., Wu Y., 2024, @doi [ ] 10.1093/mnras/stad3931 , https://ui.adsabs.harvard.edu/abs/2024MNRAS.52711206I 527, 11206

    2024 doi

  15. [23]

    arXiv:1505.05770

    Jimenez Rezende D., Mohamed S., 2015, @doi [arXiv e-prints] 10.48550/arXiv.1505.05770 , https://ui.adsabs.harvard.edu/abs/2015arXiv150505770J p. arXiv:1505.05770

    2015 doi arXiv

  16. [24]

    Keih \"a nen E., et al., 2023, @doi [ ] 10.1051/0004-6361/202142799 , https://ui.adsabs.harvard.edu/abs/2023A&A...675A...2K 675, A2

    2023 doi

  17. [25]

    Mirocha J., 2014, @doi [ ] 10.1093/mnras/stu1193 , https://ui.adsabs.harvard.edu/abs/2014MNRAS.443.1211M 443, 1211

    2014 doi

  18. [26]

    R., 2019, @doi [ ] 10.1093/mnras/sty3260 , https://ui.adsabs.harvard.edu/abs/2019MNRAS.483.1980M 483, 1980

    Mirocha J., Furlanetto S. R., 2019, @doi [ ] 10.1093/mnras/sty3260 , https://ui.adsabs.harvard.edu/abs/2019MNRAS.483.1980M 483, 1980

    2019 doi

  19. [27]

    Mirocha J., Anderson C., Chang T.-C., Dor \'e O., Lidz A., 2025, @doi [ ] 10.3847/1538-4357/adbdce , https://ui.adsabs.harvard.edu/abs/2025ApJ...983...54M 983, 54

    2025 doi

  20. [28]

    Mittal S., Kulkarni G., 2022, @doi [ ] 10.1093/mnras/stac1961 , https://ui.adsabs.harvard.edu/abs/2022MNRAS.515.2901M 515, 2901

    2022 doi

  21. [29]

    A., et al., 2024a, @doi [ ] 10.1093/mnras/stae1138 , https://ui.adsabs.harvard.edu/abs/2024MNRAS.530.4125M 530, 4125

    Monsalve R. A., et al., 2024a, @doi [ ] 10.1093/mnras/stae1138 , https://ui.adsabs.harvard.edu/abs/2024MNRAS.530.4125M 530, 4125

    doi

  22. [30]

    A., et al., 2024b, @doi [ ] 10.3847/1538-4357/ad0f1b , https://ui.adsabs.harvard.edu/abs/2024ApJ...961...56M 961, 56

    Monsalve R. A., et al., 2024b, @doi [ ] 10.3847/1538-4357/ad0f1b , https://ui.adsabs.harvard.edu/abs/2024ApJ...961...56M 961, 56

    doi

  23. [31]

    J., Bowman J

    Mozdzen T. J., Bowman J. D., Monsalve R. A., Rogers A. E. E., 2016, @doi [ ] 10.1093/mnras/stv2601 , https://ui.adsabs.harvard.edu/abs/2016MNRAS.455.3890M 455, 3890

    2016 doi

  24. [32]

    J., Bowman J

    Mozdzen T. J., Bowman J. D., Monsalve R. A., Rogers A. E. E., 2017, @doi [ ] 10.1093/mnras/stw2696 , https://ui.adsabs.harvard.edu/abs/2017MNRAS.464.4995M 464, 4995

    2017 doi

  25. [33]

    J., Mahesh N., Monsalve R

    Mozdzen T. J., Mahesh N., Monsalve R. A., Rogers A. E. E., Bowman J. D., 2019, @doi [ ] 10.1093/mnras/sty3410 , https://ui.adsabs.harvard.edu/abs/2019MNRAS.483.4411M 483, 4411

    2019 doi

  26. [34]

    B., Loeb A., 2018, @doi [ ] 10.1038/s41586-018-0151-x , https://ui.adsabs.harvard.edu/abs/2018Natur.557..684M 557, 684

    Mu \ n oz J. B., Loeb A., 2018, @doi [ ] 10.1038/s41586-018-0151-x , https://ui.adsabs.harvard.edu/abs/2018Natur.557..684M 557, 684

    2018 doi

  27. [35]

    J., et al., 2021, Experimental Astronomy, https://ui.adsabs.harvard.edu/abs/2021arXiv210401756N 51, 193

    Nambissan T. J., et al., 2021, Experimental Astronomy, https://ui.adsabs.harvard.edu/abs/2021arXiv210401756N 51, 193

    2021

  28. [36]

    Pagano M., Sims P., Liu A., Anstey D., Handley W., de Lera Acedo E., 2024, @doi [ ] 10.1093/mnras/stad3392 , https://ui.adsabs.harvard.edu/abs/2024MNRAS.527.5649P 527, 5649

    2024 doi

  29. [37]

    Philip L., et al., 2019, @doi [J. Astron. Instrum.] 10.1142/S2251171719500041 , https://ui.adsabs.harvard.edu/abs/2019JAI.....850004P 8, 1950004

    2019 doi

  30. [38]

    T., Urbano A., 2018, @doi [ ] 10.1103/PhysRevLett.121.031103 , https://ui.adsabs.harvard.edu/abs/2018PhRvL.121c1103P 121, 031103

    Pospelov M., Pradler J., Ruderman J. T., Urbano A., 2018, @doi [ ] 10.1103/PhysRevLett.121.031103 , https://ui.adsabs.harvard.edu/abs/2018PhRvL.121c1103P 121, 031103

    2018 doi

  31. [39]

    C., et al., 2018, @doi [ ] 10.1093/mnras/sty1244 , https://ui.adsabs.harvard.edu/abs/2018MNRAS.478.4193P 478, 4193

    Price D. C., et al., 2018, @doi [ ] 10.1093/mnras/sty1244 , https://ui.adsabs.harvard.edu/abs/2018MNRAS.478.4193P 478, 4193

    2018 doi

  32. [40]

    R., Loeb A., 2012, @doi [Rep

    Pritchard J. R., Loeb A., 2012, @doi [Rep. Prog. Phys.] 10.1088/0034-4885/75/8/086901 , https://ui.adsabs.harvard.edu/abs/2012RPPh...75h6901P 75, 086901

    2012 doi

  33. [41]

    O., 2020, @doi [ ] 10.3847/1538-4357/ab9b29 , https://ui.adsabs.harvard.edu/abs/2020ApJ...897..174R 897, 174

    Rapetti D., Tauscher K., Mirocha J., Burns J. O., 2020, @doi [ ] 10.3847/1538-4357/ab9b29 , https://ui.adsabs.harvard.edu/abs/2020ApJ...897..174R 897, 174

    2020 doi

  34. [42]

    J., Bigot-Sazy M.-A., Ghosh T., 2015, @doi [MNRAS] 10.1093/mnras/stv1274 , 451, 4311

    Remazeilles M., Dickinson C., Banday A. J., Bigot-Sazy M.-A., Ghosh T., 2015, @doi [MNRAS] 10.1093/mnras/stv1274 , 451, 4311

    2015 doi

  35. [43]

    Sathyanarayana Rao M., et al., 2023, @doi [Experimental Astronomy] 10.1007/s10686-023-09909-5 , https://ui.adsabs.harvard.edu/abs/2023ExA....56..741S 56, 741

    2023 doi

  36. [44]

    D., de Lera Acedo E., Handley W., Koopmans L

    Saxena A., Meerburg P. D., de Lera Acedo E., Handley W., Koopmans L. V. E., 2023, @doi [ ] 10.1093/mnras/stad1047 , https://ui.adsabs.harvard.edu/abs/2023MNRAS.522.1022S 522, 1022

    2023 doi

  37. [45]

    A., Windhorst R

    Shaver P. A., Windhorst R. A., Madau P., de Bruyn A. G., 1999, @doi [ ] 10.48550/arXiv.astro-ph/9901320 , https://ui.adsabs.harvard.edu/abs/1999A&A...345..380S 345, 380

    1999 doi arXiv

  38. [46]

    R., Sigurdson K., Pen U.-L., Stebbins A., Sitwell M., 2014, @doi [ ] 10.1088/0004-637X/781/2/57 , https://ui.adsabs.harvard.edu/abs/2014ApJ...781...57S 781, 57

    Shaw J. R., Sigurdson K., Pen U.-L., Stebbins A., Sitwell M., 2014, @doi [ ] 10.1088/0004-637X/781/2/57 , https://ui.adsabs.harvard.edu/abs/2014ApJ...781...57S 781, 57

    2014 doi

  39. [47]

    Shi Y., Deng F., Xu Y., Wu F., Yan Q., Chen X., 2022, @doi [ ] 10.3847/1538-4357/ac5965 , https://ui.adsabs.harvard.edu/abs/2022ApJ...929...32S 929, 32

    2022 doi

  40. [48]

    H., Pober J

    Sims P. H., Pober J. C., 2020, @doi [ ] 10.1093/mnras/stz3388 , https://ui.adsabs.harvard.edu/abs/2020MNRAS.492...22S 492, 22

    2020 doi

  41. [49]

    H., et al., 2023, @doi [MNRAS] 10.1093/mnras/stad610 , 521, 3273

    Sims P. H., et al., 2023, @doi [MNRAS] 10.1093/mnras/stad610 , 521, 3273

    2023 doi

  42. [50]

    Singh S., Subrahmanyan R., 2019, @doi [ApJ] 10.3847/1538-4357/ab2879 , 880, 26

    2019 doi

  43. [51]

    Astron.] 10.1038/s41550-022-01610-5 , https://ui.adsabs.harvard.edu/abs/2022NatAs...6..607S 6, 607

    Singh S., et al., 2022, @doi [Nat. Astron.] 10.1038/s41550-022-01610-5 , https://ui.adsabs.harvard.edu/abs/2022NatAs...6..607S 6, 607

    2022 doi

  44. [52]

    V., eds, American Institute of Physics Conference Series Vol

    Skilling J., 2004, in Fischer R., Preuss R., Toussaint U. V., eds, American Institute of Physics Conference Series Vol. 735, Bayesian Inference and Maximum Entropy Methods in Science and Engineering: 24th International Workshop on Bayesian Inference and Maximum Entropy Methods...

    2004 doi

  45. [53]

    Slosar A., 2017, @doi [ ] 10.1103/PhysRevLett.118.151301 , https://ui.adsabs.harvard.edu/abs/2017PhRvL.118o1301S 118, 151301

    2017 doi

  46. [54]

    J., Fialkov A., Garsden H., 2022, @doi [ ] 10.1093/mnras/stac1804 , https://ui.adsabs.harvard.edu/abs/2022MNRAS.515.1580S 515, 1580

    Spinelli M., Kyriakou G., Bernardi G., Bolli P., Greenhill L. J., Fialkov A., Garsden H., 2022, @doi [ ] 10.1093/mnras/stac1804 , https://ui.adsabs.harvard.edu/abs/2022MNRAS.515.1580S 515, 1580

    2022 doi

  47. [55]

    O., Switzer E., 2018, @doi [ ] 10.3847/1538-4357/aaa41f , https://ui.adsabs.harvard.edu/abs/2018ApJ...853..187T 853, 187

    Tauscher K., Rapetti D., Burns J. O., Switzer E., 2018, @doi [ ] 10.3847/1538-4357/aaa41f , https://ui.adsabs.harvard.edu/abs/2018ApJ...853..187T 853, 187

    2018 doi

  48. [56]

    O., 2020a, @doi [ ] 10.3847/1538-4357/ab9a3f , https://ui.adsabs.harvard.edu/abs/2020ApJ...897..132T 897, 132

    Tauscher K., Rapetti D., Burns J. O., 2020a, @doi [ ] 10.3847/1538-4357/ab9a3f , https://ui.adsabs.harvard.edu/abs/2020ApJ...897..132T 897, 132

    doi

  49. [57]

    O., 2020b, @doi [ ] 10.3847/1538-4357/ab9b2a , https://ui.adsabs.harvard.edu/abs/2020ApJ...897..175T 897, 175

    Tauscher K., Rapetti D., Burns J. O., 2020b, @doi [ ] 10.3847/1538-4357/ab9b2a , https://ui.adsabs.harvard.edu/abs/2020ApJ...897..175T 897, 175

    doi

  50. [58]

    Tegmark M., 1997, @doi [ ] 10.1086/310631 , https://ui.adsabs.harvard.edu/abs/1997ApJ...480L..87T 480, L87

    1997 doi

  51. [59]

    D., G \'o rski K

    Wandelt B. D., G \'o rski K. M., 2001, @doi [ ] 10.1103/PhysRevD.63.123002 , https://ui.adsabs.harvard.edu/abs/2001PhRvD..63l3002W 63, 123002

    2001 doi

  52. [60]

    J., Irfan M

    Wilensky M. J., Irfan M. O., Bull P., 2025, @doi [ ] 10.1093/mnras/staf668 , https://ui.adsabs.harvard.edu/abs/2025MNRAS.539.3122W 539, 3122

    2025 doi

  53. [61]

    A., 1952, @doi [ ] 10.1086/106661 , https://ui.adsabs.harvard.edu/abs/1952AJ.....57R..31W 57, 31

    Wouthuysen S. A., 1952, @doi [ ] 10.1086/106661 , https://ui.adsabs.harvard.edu/abs/1952AJ.....57R..31W 57, 31

    1952 doi

  54. [62]

    Astron.] 10.1038/s41550-022-01709-9 , https://ui.adsabs.harvard.edu/abs/2022NatAs...6..984D 6, 984

    de Lera Acedo E., et al., 2022, @doi [Nat. Astron.] 10.1038/s41550-022-01709-9 , https://ui.adsabs.harvard.edu/abs/2022NatAs...6..984D 6, 984

    2022 doi

  55. [63]

    de Oliveira-Costa A., Tegmark M., 2006, @doi [ ] 10.1103/PhysRevD.74.023005 , https://ui.adsabs.harvard.edu/abs/2006PhRvD..74b3005D 74, 023005

    2006 doi

  56. [64]

    M., Jonas J., Landecker T

    de Oliveira-Costa A., Tegmark M., Gaensler B. M., Jonas J., Landecker T. L., Reich P., 2008, @doi [ ] 10.1111/j.1365-2966.2008.13376.x , https://ui.adsabs.harvard.edu/abs/2008MNRAS.388..247D 388, 247

    2008

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.