Pith. sign in

REVIEW 3 major objections 8 minor 21 references

Studying the Temporal Variation of the Cosmic-Ray Sun Shadow Using IceCube Data

T0 review · 3 major / 8 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 cosmic-ray shadow of the Sun, measured by IceCube at median energies around 50-60 TeV, varies over time and tracks the 11-year solar activity cycle as traced by sunspot number and solar magnetic field models.

desk verdict The IceCube Sun shadow paper is a credible incremental result—a 7-year, 50-60 TeV shadow variation tracking the solar cycle with a stable Moon control—but the off-source background needs a cleaner test before the nominal significances are taken at face value. read the letter →

arxiv 1908.10148 v1 pith:6XLVIWHH submitted 2019-08-27 astro-ph.HE astro-ph.IMastro-ph.SR

Frederik Tenholt , Julia Becker Tjus , Paolo Desiati (for the IceCube Collaboration) This is my paper · ORCID
classification astro-ph.HEastro-ph.IMastro-ph.SR
keywords cosmic-raySunshadowIceCubesolarmagneticfieldactivitycyclesunspotnumberMoonMonteCarloback-trackingcoronalmodels
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 aims to show that the cosmic-ray shadow of the Sun, recorded by IceCube at a median primary energy of roughly 50-60 TeV, is not a static feature but changes over time in step with the 11-year solar activity cycle. The key evidence is that the shadow's relative deficit decreases with the average sunspot number, with a linear decline favored over a constant at 6.4 sigma, and that a simple opaque solar disk is rejected at 7.3 sigma. If true, the Sun shadow becomes a remote, time-resolved probe of the solar magnetic field near the Sun, a region that cannot be probed directly. The Moon shadow, measured with the same pipeline, stays constant, which argues that the Sun shadow variation is physical rather than detector drift.

What carries the argument

The quantity carrying the analysis is the relative deficit $RD(1^\circ) = 100\% \cdot \frac{N_{\rm on} - \langle N_{\rm off}\rangle}{\langle N_{\rm off}\rangle}$, computed inside a 1-degree radius circle around the shadow's center of gravity, with the background $\langle N_{\rm off}\rangle$ estimated from eight nearby off-source windows. Year by year this deficit is compared against three expectations: a simple opaque solar disk, the predictions obtained by back-tracking simulated cosmic-ray primaries through time-evolving PFSS and CSSS coronal magnetic field models with a per-particle passing probability, and the average sunspot number. The Moon shadow, processed identically, provides the detector-stability control.

What would settle it

Re-run the full pipeline with event directions scrambled in right ascension (or with a different set of off-source windows) and check whether a sunspot-correlated 'shadow' still appears; if it does, the background estimate is biased and the claim fails. A simpler falsifier is a full solar cycle in which the Sun shadow stays constant while sunspot number changes, which would contradict the reported 6.4 sigma linear trend.

Watch

Extended reading notes

Core claim

Using seven years of IceCube data, from May 2010 to May 2017, the paper reports that the relative deficit of cosmic rays within one degree of the Sun's shadow center is significantly lower when the Sun is active: the data favor a decreasing linear relation between the deficit and the average sunspot number over a constant relation at 6.4 sigma, and the constant solar-disk expectation is excluded at 7.3 sigma. Simulations in which primary cosmic rays are back-tracked through two time-dependent models of the coronal magnetic field (the PFSS and CSSS models) reproduce the time variation better than a disk, though with residual tensions near 3 sigma. Over the same interval the Moon shadow matches a simple lunar disk with a p-value of 0.32, indicating that the detector's angular reconstruction and event selection were stable enough for the solar variation to be taken at face value.

Load-bearing premise

The analysis assumes that the eight off-source windows provide an unbiased estimate of the background in the on-source window, so the relative deficit is purely the shadow signal; if cosmic-ray anisotropy, seasonal muon-rate changes, or detector acceptance differ between the windows, every deficit and significance is biased.

Editorial extensions

If this is right

  • The Sun shadow can serve as an indirect, energy-resolved measurement of the solar magnetic field close to the Sun, where direct measurements are not available.
  • A longer IceCube data span covering more of the solar cycle can test whether the linear sunspot-deficit relation persists through solar minimum and into the next maximum.
  • The residual ~3 sigma tension with both field models motivates including coronal mass ejections, a more realistic solar wind profile, and corrected magnetogram scales in future simulations.
  • Because the Moon shadow is stable over the same seven years, the Sun shadow variation cannot be attributed to detector acceptance or reconstruction drift.
  • At these energies the shadow responds to solar activity, so combining with lower-energy and higher-rigidity cosmic-ray observatories can map the rigidity dependence of the effect.

Reading between the lines

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

  • A natural next test, not performed in the paper, is to apply the same background-subtracted deficit to individual solar rotations to search for transient shadow changes associated with coronal mass ejections.
  • The reported anti-correlation suggests the solar magnetic field acts as an energy-dependent scattering opacity; applying the same analysis at lower primary energies could reveal whether the deficit saturates at solar maximum.
  • If the off-source background estimate is independently validated with synthetic or scrambled events, the deficit-sunspot relation could become a ground-based constraint on heliospheric magnetic field models, complementing in-situ spacecraft data.
  • The ~3 sigma residual disagreement with both PFSS and CSSS models points toward missing magnetic structure such as non-potential fields; a time-dependent non-potential simulation would be the direct test.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 8 minor

Summary. This proceedings paper (ICRC 2019) reports a seven-year (2010-2017) measurement of the cosmic-ray Moon and Sun shadows with IceCube at a median primary energy around 50-60 TeV. The Moon shadow is used as a control and is consistent with the expectation from a geometric lunar disk (p = 0.32). For the Sun, the relative deficit inside a 1-degree-radius circle around the center of gravity of the shadow, Eq. (2.3), falls from roughly 6% in the 2010/11 season to about 2% in 2016/17. The authors report a 7.3-sigma rejection of a simple solar-disk model, a 6.4-sigma preference for a decreasing linear relation between the relative deficit and the average sunspot number, and agreement with PFSS/CSSS solar magnetic field models that is much better than the disk expectation, with residual tensions of 3.0 sigma (PFSS) and 2.8 sigma (CSSS). The paper concludes that the TeV Sun shadow varies over the 11-year solar cycle, consistent with expectations from the solar magnetic field cycle.

Significance. If correct, the central result—a solar-cycle-correlated variation of the cosmic-ray Sun shadow at 50-60 TeV, an energy regime above the Tibet and ARGO measurements—is a noteworthy probe of the coronal and interplanetary magnetic field. The analysis has genuine strengths: the correlation is established against external WDC-SILSO sunspot numbers and is not fitted; the PFSS/CSSS predictions are driven by SOLIS magnetograms rather than by the shadow data; the stable Moon shadow (Fig. 2) controls for detector drift, pointing, and acceptance over the same seven years; and the concluding section explicitly acknowledges the factor-of-two magnetogram scale uncertainty and the neglect of CMEs. The weaknesses are statistical rather than conceptual: the off-source background windows are not demonstrated to be free of solar modulation, the quoted significances are Poisson-only with no propagated detector or model systematics, and the test statistics are not fully specified. These issues are identifiable and fixable within the manuscript's scope, and none of them constitutes an internal inconsistency in the analysis.

major comments (3)
  1. [§2.2, Eq. (2.3), Figs. 3-4] The 6.4-sigma trend in Fig. 4 and every annual RD(1°) point in Fig. 3 rest on the assumption that <Noff> in Eq. (2.2), the average of the eight off-source windows, is an unbiased, Sun-free background. The manuscript provides no test of this assumption: the angular separation of the off-source windows from the Sun's nominal position is not stated, and the §3 simulations (which produce angular structure over the 3-degree by 3-degree maps shown in §4) are not used to estimate the solar modulation at the off-source positions. Because the magnetic deflection pattern and its magnitude vary with the solar cycle, any solar-cycle-dependent component in <Noff> would bias the yearly points in a way that correlates with sunspot number, and the Poisson-only significance would not capture it. The Moon control in Fig. 2 cannot rule this out because the Moon has no magnetic field. The contamination may well be negligible at the actual offsets, but the paper should demonstrate that, for example by reporting the year-by-year stability of the eight individual off-source windows, by estimating the fractional modulation at the off-source positions with the back-tracking simulations, or by recomputing RD(1°) with alternative off-source geometries.
  2. [§4 (Figs. 3-4), §5] The headline significances are not reproducible as stated. The paper does not give the test statistic (chi-square, likelihood ratio, or other), the number of degrees of freedom, or the way the error bars on the annual RD(1°) points are computed for (i) the 7.3-sigma rejection of the solar-disk model, (ii) the 6.4-sigma preference for a decreasing linear relation versus a constant, and (iii) the 3.0-sigma (PFSS) and 2.8-sigma (CSSS) model deviations. All quoted significances are Poisson-only; no systematic uncertainty is propagated. This matters most for the model comparison, where the factor-of-two magnetogram scale uncertainty acknowledged in §5 via ref [21] is of the same order as the 3.0/2.8-sigma tensions; the claim that the models agree with the data better than the disk should be re-quantified with a systematic band on the model predictions. In addition, if any of the p-values are computed from the smoothed 0.1-degree-binned maps, the 0.7-degree boxcar smoothing makes neighboring bins strongly correlated, and treating them as independent would overstate the significance; the scale uncertainty in ref [21] needs to be applied to the model predictions, not only cited.
  3. [§2.2] The center-of-gravity correction is a selection on the same data that are then used to measure RD(1°): the shadow center is defined by the bins with smoothed deficit greater than 3%, and the deficit is then measured around that center. Self-centering biases the recovered deficit upward, and the bias is expected to be largest for the shallow, noisier shadows of the solar-maximum years, which are precisely the points that set the slope in Fig. 4. Please quantify the bias with pseudo-experiments (for example, injecting shadow templates of known depth and center shift into the event sample and repeating the COG correction), or demonstrate through a comparable cross-check that the solar-cycle trend in Figs. 3-4 is unchanged. The paper should also state whether there are any years in which no bins pass the 3% threshold and how such years are handled.
minor comments (8)
  1. [§2.2, Eqs. (2.2)-(2.3)] As defined, the 'relative deficit' is negative for a shadow, while Figs. 2-4 plot positive values; please state explicitly that the plotted quantity is the absolute value of RD(1°).
  2. [Fig. 1, §2.2] Please state the angular separation between the on-source window and the nearest off-source windows; the figure suggests a 3x3 arrangement of touching 6-degree by 6-degree boxes, implying roughly 3 degrees between the on-source center and the nearest off-source edges.
  3. [§3, §4] The text says the simulations are processed like the data; please confirm explicitly that the simulated model predictions in Fig. 3 undergo the identical center-of-gravity correction and 1-degree-circle RD definition, since the comparison in that figure depends on it.
  4. [Fig. 4] Specify the time interval over which the average sunspot number is computed for each season, given that the Sun filter collects data for roughly 90 days per season (§2.1).
  5. [§1.2] 'Down-going muons, with an energy of roughly more than 400 GeV' should be rephrased; also, give the median energy of the Sun sample rather than only the range quoted.
  6. [§3] State how many statistically independent simulated primaries remain after the resampling (duplicating each event by factors of 20-100 does not add statistical power), and note that the Delta-delta distribution is preserved, not resampled.
  7. [§5, abstract] 'These numbers, independent on the exact solar field model prove a variation' should read 'independent of' and a verb such as 'indicate' or 'support' is more appropriate than 'prove' for a Poisson-only analysis; the abstract statement that the results correlate 'with theoretical models of the solar magnetic field' is also stronger than the 3.0/2.8-sigma tensions justify.
  8. [References] Ref. [15] is cited in its arXiv preprint form although it provides the back-tracking framework used here; please update to the published version if available.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity found: the Sun-shadow time variation is measured from IceCube event counts and compared with external sunspot data and magnetogram-driven PFSS/CSSS models.

full rationale

The derivation chain is data-driven rather than self-referential. The relative deficit RD(1°) is defined in Eq. (2.3) directly from on-source and eight off-source event counts, with no fitted parameter entering the definition. The temporal variation shown in Fig. 3 is an unbinned comparison of these counts across seasons, and the sunspot correlation in Fig. 4 uses external WDC-SILSO sunspot numbers as the independent variable; the 6.4σ preference for a decreasing linear relation is a statistical fit to the measured RD values, not a parameter fitted from the shadow. The PFSS/CSSS predictions are magnetogram-driven forward models cited to [15], and the paper explicitly reports residual tensions (3.0σ and 2.8σ) rather than tuning those models to match the data. The only self-citation, [15], supplies the back-tracking technique but is not used to define the measured quantity, and its predictions are externally falsifiable and partially disfavored by the data. No equation reduces by construction to its inputs, and no fitted quantity is renamed as a prediction. Possible off-source-window bias is a correctness risk, not a circularity, because it concerns the validity of the background estimate rather than a formal reduction of the result to its assumptions.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central claim introduces no new entities. The analysis depends on solar magnetic field models and IceCube simulations from prior work; the main free choices are analysis parameters. The solar-cycle variation claim does not reduce to a fitted parameter, but the statistical significance is computed under nominal Poisson errors without propagated systematic uncertainties.

free parameters (4)
  • Quality cut thresholds (sigma, rlogl) = sigma < 0.71 deg, rlogl < 8.1
    Optimized in Section 2.1 to maximize expected shadow significance; changes the sample and therefore the measured deficit.
  • Smoothing radius = 0.7 deg
    Chosen in Section 2.2 to approximately equal the median angular resolution; affects the center-of-gravity estimate.
  • Center-of-gravity deficit threshold = 3% relative deficit
    In Section 2.2 only smoothed bins above 3% define the shadow center, a selection that can bias the final corrected deficit.
  • Search radius for deficit = 1.0 deg
    Chosen in Section 2.2 as a trade-off between capturing shadowed events and including background; uncertainty not propagated.
assumptions (5)
  • domain assumption Multi-TeV muon arrival directions approximate the primary cosmic-ray direction to within about 0.1 degrees.
    Invoked in Section 1.2 to justify treating reconstructed muon directions as a shadow map; based on ref [6].
  • domain assumption The eight off-source windows provide an unbiased estimate of the background in the on-source window.
    Used in Eq. (2.2) and Fig. 1; any anisotropy or acceptance difference between windows biases all relative deficits.
  • domain assumption Event counts in the bins follow Poisson statistics.
    Basis for the quoted significances in Section 2.1 and Section 4; resampling and smoothing introduce correlations that are not folded into the errors.
  • domain assumption PFSS and CSSS potential field models plus an analytic Parker spiral adequately represent the solar magnetic field encountered by TeV cosmic rays.
    Used in Section 3 to compute passing probabilities; Section 5 acknowledges magnetogram scale errors of about a factor of two, neglect of CMEs, and the potential-field approximation.
  • domain assumption The CORSIKA simulations with Hillas-Gaisser weights and the resampling procedure preserve the distribution of reconstructed-minus-primary directions.
    Used in Section 3 to build expected shadow templates; the small original simulation sample and the 20-100 fold duplication may not capture rare large deflections.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Studying the Temporal Variation of the Cosmic-Ray Sun Shadow Using IceCube Data." pith.science (2026). https://pith.science/paper/6XLVIWHH

@misc{pith2026190810148,
  author       = {Pith},
  title        = {Pith review of: Studying the Temporal Variation of the Cosmic-Ray Sun Shadow Using IceCube Data},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6XLVIWHH}},
  note         = {Machine review of arXiv:1908.10148}
}
abstract

The shadowing effect of the Moon and Sun in TeV cosmic rays has been measured with high statistical significance by several experiments. Unlike particles from directions close to the Moon, however, charged particles passing by the neighborhood of the Sun are affected not only by the geomagnetic but also by the solar near- and interplanetary-magnetic field. Since the latter undergoes a well-known 11-year cycle -- during which it can become highly disordered -- the cosmic-ray shadow cast by the Sun as observed on Earth is expected to change over time. We present an update of the analysis of the cosmic-ray Moon and Sun shadows using data taken with the IceCube Neutrino Observatory. With a median energy after quality cuts of approximately $50-60\,$TeV, depending on the cosmic-ray flux model used, primary cosmic rays inducing events which pass IceCube's Sun shadow filter have a comparatively high energy. While the results for the Moon shadow confirm the stability of the IceCube observatory, the results for the Sun shadow exhibit a clear variation correlating with solar activity and theoretical models of the solar magnetic field.

Figures

Figures reproduced from arXiv: 1908.10148 by the authors.

Figure 1
Figure 1. Layout of the analysis with 8 off-source bins for background determination (red) and one on￾source bin (black/grey). The dashed boxes represent the region that is shown in the plots in Section 4. off-source windows with a size of 6◦ ×6 ◦ are defined. A schematic view of the geometry of these windows is given in [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Relative deficit of the Moon as a function of time. 6 [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Relative deficit of the Sun as a function of time. Magnetogram input data for both models were acquired by SOLIS instruments operated by NISP/NSO/AURA/NSF. The results for the relative deficit according to Eq. 2.3 for the Moon as a function of time are shown in [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: shows the relative deficit as a function of the average sunspot number. The data favor a 0 20 40 60 80 100 120 140 Average Sunspot Number 7 6 5 4 3 2 1 R elativ e D eficit (1.0°) in % Linear Regression This work/WDC-SILSO [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

21 extracted references · 19 canonical work pages

  1. [21]

    Riley et al., Sol

    P. Riley et al., Sol. Phys. 289 (Mar, 2014) 769–792. 8

    work page 2014

  2. [1]

    Amenomori et al., Phys

    M. Amenomori et al., Phys. Rev. Lett. 111 (July, 2013) 011101

    work page 2013

  3. [2]

    Amenomori et al., Phys

    M. Amenomori et al., Phys. Rev. Lett. 120 (Jan., 2018) 031101

    work page 2018

  4. [3]

    Amenomori et al., ApJ 860 (June, 2018) 13

    M. Amenomori et al., ApJ 860 (June, 2018) 13

    work page 2018

  5. [4]

    Chen and Y

    S. Chen and Y . Nan, PoS ICRC2017 (2018)

    work page 2018

  6. [5]

    M. G. Aartsen et al., JINST 12 (2017) P03012

    work page 2017

  7. [6]

    M. G. Aartsen et al., Phys. Rev. D 89 (May, 2014) 102004

    work page 2014

  8. [7]

    M. G. Aartsen et al., ApJ 826 (Aug., 2016) 220

    work page 2016

Show all 21 references
  1. [8]

    A. U. Abeysekara et al., arXiv:1812.05682

    arXiv

  2. [9]

    M. G. Aartsen et al., Phys. Rev. D 88 (2013) 042004

    2013

  3. [10]

    M. P. K. Andeen, Ultra High Energy Cosmic Ray Conference, Paris, France (2019)

    2019

  4. [11]

    M. G. Aartsen et al., ApJ 872 (Feb., 2019) 133

    2019

  5. [12]

    Abbasi et al., ApJ 732 (Apr., 2011) 18

    R. Abbasi et al., ApJ 732 (Apr., 2011) 18

    2011

  6. [13]

    Heck et al., F orschungszentrum Karlsruhe Report FZKA6019 (1998) 90

    D. Heck et al., F orschungszentrum Karlsruhe Report FZKA6019 (1998) 90

    1998

  7. [14]

    T. K. Gaisser, Astropart. Phys. 35 (July, 2012) 801–806

    2012

  8. [15]

    Becker Tjus et al., arXiv:1903.12638

    J. Becker Tjus et al., arXiv:1903.12638

    1903 arXiv

  9. [16]

    M. D. Altschuler and G. Newkirk, Sol. Phys. 9 (Sept., 1969) 131–149

    1969

  10. [17]

    K. H. Schatten, J. M. Wilcox, and N. F. Ness, Sol. Phys. 6 (Mar., 1969) 442–455

    1969

  11. [18]

    Zhao and J

    X. Zhao and J. T. Hoeksema, Adv. Space Res. 16 (Aug., 1995)

    1995

  12. [19]

    C. N. Arge and V . J. Pizzo, J. Geophys. Res. 105 (May, 2000) 10465–10480

    2000

  13. [20]

    SILSO World Data Center, International Sunspot Number Monthly Bulletin and online catalogue (2010-2017)

    2010

Pith tools

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