Pith. sign in

REVIEW 2 major objections 6 minor 25 references

Is the lack of power anomaly in the CMB correlated with the orientation of the Galactic plane?

T0 review · 2 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read The paper claims that the CMB's high-latitude low-variance (lack-of-power) anomaly survives random rotations of the Galactic orientation, at about 2.8 to 3.1 sigma, so it is not a mask-selection fluke.

desk verdict A careful MC study of the CMB low-variance anomaly that adds a rotation look-elsewhere test; the central qualitative claim holds, but the headline 2.8–3.1 sigma are quoted at the most extreme mask without a global mask/estimator correction. read the letter →

arxiv 1908.10637 v2 pith:JWEDJUFE submitted 2019-08-28 astro-ph.CO

U.Natale , A.Gruppuso , D.Molinari , P.Natoli This is my paper · ORCID
classification astro-ph.CO
keywords cosmicmicrowavebackgroundCMBanomalieslackofpowerlowvarianceGalacticmasklook-elsewhereeffectangularspectrumPlanck2018
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 asks whether the CMB's 'lack of power' anomaly—the observation that large-scale temperature fluctuations are weaker than the standard cosmological model predicts—is an artifact of choosing the orientation of our Galaxy as a reference. It reproduces the known trend that the anomaly becomes stronger as more of the Galactic plane is masked out, reaching about 3 sigma in the most aggressive high-latitude mask. The new step is to rotate data and simulated maps randomly, which mimics a look-elsewhere effect across all possible orientations of the Galaxy. The claim is that the high-latitude low variance remains anomalous after this resampling: the lower-tail probability is 0.5 percent for one estimator and 0.2 percent for the trend estimator, corresponding to roughly 2.8 and 3.1 sigma. If correct, the orientation of our Galaxy is not what is producing the anomaly, and explanations must account for why the deficit is strongest away from the Galactic plane.

What carries the argument

The central object is the map variance $V = \sum_{\ell=2}^{29} \frac{2\ell+1}{4\pi} C_\ell$, estimated with the Quadratic Maximum Likelihood angular power-spectrum estimator. The new machinery is rotation-resampling: each map is rotated many times via Wigner rotation matrices applied to spherical-harmonic coefficients, and the distribution of $V$ across rotations defines a lower-tail probability for each map. The $r$-estimator $r=(V_{\rm std}-V_{\rm mask})/\max_j(V_{\rm std}^{(j)}-V_{\rm mask}^{(j)})$ compares the observed variance drop between the standard and extended masks with the largest drop among rotations, separating the anomaly's dependence on the mask's sky fraction from its dependence on orientation.

What would settle it

Recompute the lower-tail probability and r-estimator using a full-sky CMB map from an independent experiment, or a map in which the Galactic-plane gap is filled by a validated inpainting scheme rather than masked; if the high-latitude variance is no longer in the lowest 0.5 percent of rotated Lambda CDM realisations, the claimed orientation-stable anomaly is a mask artifact.

Watch

Extended reading notes

Core claim

We show that the lack of power at high Galactic latitude is substantially stable against the look-elsewhere effect induced by random rotations of the Galactic orientation. The variance of the temperature map, computed from the angular power spectrum up to l=29, decreases as more sky around the Galactic plane is masked; random rotations of the same maps show that the observed decrease is near the extreme of what rotations can produce, with r around 0.88-0.90 and an upper-tail probability of 0.2 percent. The lower-tail probability estimator gives 0.5 percent for Commander and 1.3 percent for SMICA in the most aggressive mask, meaning only 5 (or 13) of 1000 Lambda CDM realisations have a more anomalous orientation. Repeating the analysis with simulations constrained to have the observed low variance gives essentially the same result, so the anomaly is not merely a consequence of the map having low power to begin with.

Load-bearing premise

The whole significance calculation assumes that the standard cosmological model fitted to the Planck data is the right description of the true large-scale sky; if that fit already absorbed part of the missing power, the quoted probabilities would shift.

Editorial extensions

If this is right

  • The high-latitude low-variance anomaly is not a selection effect of Galactic orientation; any explanation must produce a deficit that is intrinsically stronger away from the Galactic plane.
  • The decreasing variance trend with mask size is itself anomalous against rotations, with the observed maps falling in the upper 0.2 percent of the r-estimator distribution, so the anisotropy of the anomaly is part of the signal.
  • The anomaly persists when simulations are constrained to have the same low total variance as the data, ruling out the interpretation that a generally low-variance sky naturally produces this latitude trend.
  • The effect is dominated by the quadrupole and octupole, and at low latitudes by odd multipoles, pointing to the lowest multipoles as the locus of the signal.

Reading between the lines

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

  • A natural testable extension is to apply the same rotation-resampling to other CMB anomalies, such as hemispherical asymmetry or parity asymmetry, to see whether their significance is also orientation-stable.
  • If the anomaly is genuinely orientation-independent, then masks centred on randomly chosen sky directions should not reproduce the same high-latitude deficit unless they exclude the actual Galactic plane; this could be checked directly.
  • The result sharpens the challenge for models that try to explain the low-l power with a localized physical feature: such a feature would have to avoid the Galactic plane rather than merely suppress power globally.
  • Combining rotation-resampling with a validated full-sky inpainting method could test whether the mask itself, through mode coupling, contributes to the apparent anomaly.
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 / 6 minor

Summary. The paper investigates whether the well-known lack-of-power anomaly at large angular scales in the CMB is specifically associated with high Galactic latitude, and whether this association could be explained by the particular orientation of the Galactic plane. Using Planck 2018 Commander and SMICA temperature maps, five Galactic masks, a quadratic maximum-likelihood power-spectrum estimator, and 10^5 Lambda-CDM simulations, the authors first reproduce the known result that the variance lower-tail probability decreases as the Galactic mask grows, reaching about 0.3-0.5% for the most extended Ext30 mask. They then apply random rotations to the maps and define two estimators: the LTP estimator, which measures how often a random rotation yields a lower variance than the unrotated map, and the r estimator, which measures the variance drop relative to the largest drop across rotations. For the Ext30 mask they report a 0.5% LTP value (about 2.8 sigma) for Commander and a 0.2% upper-tail probability (about 3.1 sigma) for the r estimator. The analysis is repeated with a low-variance constrained ensemble (ensemble 1) and gives similar results, and robustness checks with different thresholds and with Planck 2015 data are provided in the appendices.

Significance. If the result holds, it is a useful and interesting contribution to the CMB-anomaly literature: it shows that the high-latitude lack of power is not simply a consequence of the Galactic plane orientation being special, and it provides a concrete rotation-based look-elsewhere procedure. The Monte Carlo design is careful and extensive: 10^5 Lambda-CDM realisations, QML C_ell estimation, two component-separation maps, five masks, a low-variance constrained ensemble, a validation of the rotation generator in Appendix A, threshold-dependence checks in Appendix C, and a comparison with the Planck 2015 release in Appendix B. The paper is honest about the exploratory nature of several choices. The main weakness is that the headline significances are quoted for the most extreme of five masks and for two estimators without a global correction for the mask and estimator scan, so the quantitative claim as stated is not yet fully supported.

major comments (2)
  1. [Section 3.1.1 and Section 3.1.2, Tables 2 and 3] The headline significances (2.8 sigma from the LTP estimator and 3.1 sigma from the r estimator) are quoted at the Ext30 mask, which is the most extreme of the five masks in Table 1 and is the mask where the monotonic trend in Figure 4 is largest. No correction is applied for scanning over the five masks or over the two estimators. Under the null, the minimum LTP across masks is stochastically smaller than the LTP at any fixed mask, so the reported 0.5% and 0.2% values are post-selection. A simple Bonferroni correction over five masks gives at least roughly 2.5% (LTP) and 1% (r), and the true global p-value after accounting for correlations among masks and estimators could be a few percent. This is load-bearing because the abstract and conclusions explicitly quote the 2.8-3.1 sigma levels as the evidence that the anomaly survives the look-elsewhere effect. Please provide a global statistic, for example the distribution over simulations of the minimum LTP (or maximum r) across masks and estimators, and report the fraction of simulations that exceed the observed value, or explicitly re-label the quoted numbers as mask-dependent exploratory findings.
  2. [Section 4 and Tables 5-6] The similarity of the ensemble 1 results is presented as supporting the main conclusion, but ensemble 1 was selected after the fact to reproduce the observed low variance. This is a conditional test, not an independent confirmation of the anomaly. Its valid and useful role is to show that a low total variance in the standard mask is not by itself sufficient to generate the high-latitude trend; however, the text should state this caveat explicitly rather than implying that the ensemble 1 analysis independently validates the 2.8-3.1 sigma result. The distinction matters because the ensemble 1 p-values inherit the same mask-selection issue identified above and are also conditioned on the data variance.
minor comments (6)
  1. [Abstract and Section 1] The word 'bizzarre' should be 'bizarre', and in Section 5 'simular' should be 'similar'.
  2. [Section 5, final paragraph] The text states that 'only 5 maps out of 10^5 have a LTP... smaller than the Planck Commander data', but Section 3.1.1 and Table 2 report 5 out of 10^3 (0.5%). This is a numerical inconsistency that should be corrected.
  3. [Figure 10 caption and Section 4] The caption of Figure 10 appears to swap 'Left panel' and 'Right panel' relative to the text: the panel showing the LTP versus sky fraction is on the left, while the panels with l_min = 3 and l_min = 4 are on the right. Please fix the caption.
  4. [Equation (3.2)] The denominator of the r-estimator is described as the maximum over rotations, but the unrotated case is included only in a footnote. The main text should explicitly state that the denominator includes the unrotated orientation, since the sign and the normalization of r depend on this convention.
  5. [Section 2.2] The null hypothesis is the Planck 2018 best-fit Lambda-CDM model, which was derived from the same data being tested. The paper should state explicitly that all p-values are conditional on this model and briefly comment on the expected direction of any bias; a one-sentence caveat in Section 2.2 or Section 5 would suffice.
  6. [Table 2] The left panel uses the notation V_c^{(rot)} < V_c, which is not defined in the table caption; please define the notation there.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the rotation look-elsewhere test is calibrated by Monte Carlo on Lambda CDM simulations, and the constrained-ensemble analysis is a conditional robustness check, not a fitted prediction.

full rationale

The paper's central claim is derived without fitting the target anomaly. Section 2.2 defines the null as 10^5 maps 'randomly extracted from the Planck 2018 best-fit model'; no parameter is adjusted to the low-variance feature. Section 3.1.1 computes for each simulation the lower-tail probability of its own rotation ensemble and compares the data's value to the resulting Monte Carlo distribution, so the rotation look-elsewhere effect is calibrated rather than assumed away. The r-estimator in Section 3.1.2 is an explicitly defined statistic normalized by the maximum decrease across rotations, and its significance is again Monte Carlo. The 'ensemble 1' analysis in Section 4 conditions on the observed low variance in the Std 2018 mask and then asks whether the mask-dependent trend persists; the conditioning variable is not the target quantity, so the result does not reduce to the input. The use of the Planck 2018 best-fit model as null is a standard caveat because the model is derived from the same data, and the self-references (e.g. Refs. [3], [15], [16], [24]) are background or code references whose relevant claims are reproduced in the present analysis. The reviewer concern about quoting the Ext30 mask without a global look-elsewhere correction over masks is a statistical multiplicity issue, not a definitional circularity.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The central claim rests on the standard cosmological null model, on the fidelity of the component-separated maps, and on the uniformity of the rotation sampling. No physical entities are introduced. The only hand-chosen analysis parameter is the ensemble 1 variance window, and its robustness is tested in Appendix C.

free parameters (1)
  • Ensemble 1 variance selection window = V_c plus or minus 20 uK^2, V_c = 2090.02 uK^2
    Hand-chosen window used to select 10^3 low-variance Lambda CDM realisations from the 10^5 set (Section 2.2). The central ensemble-0 results do not use it, and Appendix C shows stability for 10 and 30 uK^2 windows.
assumptions (3)
  • domain assumption Planck 2018 best-fit Lambda CDM model is the correct null distribution for low-l CMB temperature variance.
    Section 2.2 generates all simulations from this model; if the fitted parameters already contain the anomaly, the p-values are biased.
  • domain assumption Commander and SMICA 2018 maps are free of significant foreground or systematic residuals at large angular scales after masking.
    Section 2.1 uses the component-separated maps plus 2 uK regularization noise; a Galactic-plane-correlated residual could mimic the orientation effect.
  • domain assumption The set of 1000 random rotations covers the rotation group uniformly enough to make the LTP distribution uniform under Lambda CDM.
    Appendix A validates coverage through the C0/C1 ratio of order 10^3, but residual non-uniformity would propagate into the Monte Carlo p-values.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Is the lack of power anomaly in the CMB correlated with the orientation of the Galactic plane?." pith.science (2026). https://pith.science/paper/JWEDJUFE

@misc{pith2026190810637,
  author       = {Pith},
  title        = {Pith review of: Is the lack of power anomaly in the CMB correlated with the orientation of the Galactic plane?},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JWEDJUFE}},
  note         = {Machine review of arXiv:1908.10637}
}
abstract

The lack of power at large angular scales in the CMB temperature anisotropy pattern is a feature known to depend on the size of the Galactic mask. Not only the large scale anisotropy power in the CMB is lower than the best-fit $\Lambda$CDM model predicts, but most of the power seems to be localised close to the Galactic plane, making high-Galactic latitude regions more anomalous. We assess how likely the latter behaviour is in a $\Lambda$CDM model by extracting simulations from the {\it Planck} 2018 fiducial model. By comparing the former to {\it Planck} data in different Galactic masks, we reproduce the anomaly found in previous works, at a statistical significance of $\sim 3 \, \sigma$. This result suggests the existence of a bizzarre correlation between the particular orientation of the Galaxy and the lack of power anomaly. To test this hypothesis, we perform random rotations of the {\it Planck} 2018 data and compare these to similarly rotated $\Lambda$CDM realisations. We find that, among all possible rotations, the lower-tail probability of the observed high-Galactic latitude data variance is still low at the level of $2.8 \, \sigma$. Furthermore, the lowering trend of the variance when moving from low- to high-Galactic latitude is anomalous in the data at $\sim 3\,\sigma$ when comparing to $\Lambda$CDM rotated realisations. This shows that the lack of power at high Galactic latitude is substantially stable against the "look elsewhere" effect induced by random rotations of the Galaxy orientation. Moreover, this analysis turns out to be substantially stable if we employ, in place of generic $\Lambda$CDM simulations, a specific set whose variance is constrained to reproduce the observed data variance.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 8 canonical work pages

  1. [1]

    Monteserin, R

    C. Monteserin, R. B. B. Barreiro, P. Vielva, E. Martinez-Gonzalez, M. P. Hobson and A. N. Lasenby, Mon. Not. Roy. Astron. Soc. 387 (2008) 209 doi:10.1111/j.1365-2966.2008.13149.x [arXiv:0706.4289 [astro-ph]]

    work page arXiv 2008

  2. [3]

    Gruppuso, P

    A. Gruppuso, P. Natoli, F. Paci, F. Finelli, D. Molinari, A. De Rosa and N. Mandolesi, JCAP 1307 (2013) 047 doi:10.1088/1475-7516/2013/07/047 [arXiv:1304.5493 [astro-ph.CO]]

    arXiv 2013

  3. [4]

    Astrophys

    Planck Collaboration XXIII, Astron. Astrophys. 571 (2014) A23 doi:10.1051/0004-6361/201321534 [arXiv:1303.5083 [astro-ph.CO]]

    arXiv 2014

  4. [5]

    Astrophys

    Planck Collaboration XVI, Astron. Astrophys. 594 (2016) A16 doi:10.1051/0004-6361/201526681 [arXiv:1506.07135 [astro-ph.CO]]

    arXiv 2016

  5. [6]

    Planck Collaboration VII, arXiv:1906.02552 [astro-ph.CO]. – 21 –

    arXiv 1906

  6. [7]

    C. J. Copi, D.Huterer, D. J. Schwarz, and G. D. Starkman, Phys. Rev. D 75 (2007) doi:10.1103/PhysRevD.75.023507 [astro-ph/0605135]

    work page Pith review arXiv 2007

  7. [9]

    C. J. Copi, D. Huterer, D. J. Schwarz and G. D. Starkman, Adv. Astron. 2010 (2010) 847541 doi:10.1155/2010/847541 [arXiv:1004.5602 [astro-ph.CO]]

    arXiv 2010

  8. [10]

    D. J. Schwarz, C. J. Copi, D. Huterer and G. D. Starkman, Class. Quant. Grav. 33 (2016) no.18, 184001 doi:10.1088/0264-9381/33/18/184001 [arXiv:1510.07929 [astro-ph.CO]]

    arXiv 2016

Show all 25 references
  1. [11]

    J. Muir, S. Adhikari and D. Huterer, Phys. Rev. D 98 (2018) no.2, 023521 doi:10.1103/PhysRevD.98.023521 [arXiv:1806.02354 [astro-ph.CO]]

    2018 arXiv

  2. [12]

    Gruppuso and A

    A. Gruppuso and A. Sagnotti, Int. J. Mod. Phys. D 24 (2015) no.12, 1544008 doi:10.1142/S0218271815440083 [arXiv:1506.08093 [astro-ph.CO]]

    2015 arXiv

  3. [13]

    Gruppuso, Phys

    A. Gruppuso, Phys. Rev. D 76 (2007) 083010 doi:10.1103/PhysRevD.76.083010 [arXiv:0705.2536 [astro-ph]]

    2007 arXiv

  4. [14]

    E. F. Bunn and A. Bourdon, Phys. Rev. D 78 (2008) 123509 doi:10.1103/PhysRevD.78.123509 [arXiv:0808.0341 [astro-ph]]

    2008 arXiv

  5. [15]

    Gruppuso, N

    A. Gruppuso, N. Kitazawa, N. Mandolesi, P. Natoli and A. Sagnotti, Phys. Dark Univ. 11 (2016) 68 doi:10.1016/j.dark.2015.12.001 [arXiv:1508.00411 [astro-ph.CO]]

    2016 arXiv

  6. [16]

    Gruppuso, N

    A. Gruppuso, N. Kitazawa, M. Lattanzi, N. Mandolesi, P. Natoli and A. Sagnotti, Phys. Dark Univ. 20 (2018) 49 doi:10.1016/j.dark.2018.03.002 [arXiv:1712.03288 [astro-ph.CO]]

    2018 arXiv

  7. [17]

    Kim and P

    J. Kim and P. Naselsky, Astrophys. J. 714 (2010) L265 doi:10.1088/2041-8205/714/2/L265 [arXiv:1001.4613 [astro-ph.CO]]

    2010 arXiv

  8. [18]

    Kim and P

    J. Kim and P. Naselsky, Phys. Rev. D 82 (2010) 063002 doi:10.1103/PhysRevD.82.063002 [arXiv:1002.0148 [astro-ph.CO]]

    2010 arXiv

  9. [19]

    Gruppuso, F

    A. Gruppuso, F. Finelli, P. Natoli, F. Paci, P. Cabella, A. De Rosa and N. Mandolesi, Mon. Not. Roy. Astron. Soc. 411 (2011) 1445 doi:10.1111/j.1365-2966.2010.17773.x [arXiv:1006.1979 [astro-ph.CO]]

    2011

  10. [20]

    Astrophys

    Planck Collaboration XI, Astron. Astrophys. 594 (2016) A11 doi:10.1051/0004-6361/201526926 [arXiv:1507.02704 [astro-ph.CO]]

    2016 arXiv

  11. [21]

    Planck Collaboration IV, arXiv:1807.06208 [astro-ph.CO]

    arXiv

  12. [22]

    Planck Collaboration V, arXiv:1907.12875 [astro-ph.CO]

    1907 arXiv

  13. [23]

    K. M. Gorski, E. Hivon, A. J. Banday, B. D. Wandelt, F. K. Hansen, M. Reinecke and M. Bartelman, Astrophys. J. 622 (2005) 759 doi:10.1086/427976 [astro-ph/0409513]

    2005 arXiv

  14. [24]

    Gruppuso, A

    A. Gruppuso, A. De Rosa, P. Cabella, F. Paci, F. Finelli, P. Natoli, G. de Gasperis and N. Mandolesi, Mon. Not. Roy. Astron. Soc. 400 (2009) 463 doi:10.1111/j.1365-2966.2009.15469.x [arXiv:0904.0789 [astro-ph.CO]]

    2009

  15. [25]

    Tegmark, Max, Phys. Rev. D 55” (1997) 5895-5907 doi:10.1103/PhysRevD.55.5895 [arXiv:9611174[astro-ph]]

    1997 doi

  16. [26]

    and de Oliveira-Costa, A., Phys

    Tegmark, M. and de Oliveira-Costa, A., Phys. Rev. D 64” (2001) 063001 doi:10.1103/PhysRevD.64.063001 [arXiv:0012120[astro-ph]]

    2001 doi

  17. [27]

    Molinari, A

    D. Molinari, A. Gruppuso, G. Polenta, C. Burigana, A. De Rosa, P. Natoli, F. Finelli and F. Paci, Mon. Not. Roy. Astron. Soc. 440 (2014) no.2, 957 doi:10.1093/mnras/stu386 [arXiv:1403.1089 [astro-ph.CO]]. – 22 –

    2014 arXiv

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.