REVIEW 4 minor 12 references
The Milky Way's Galactic plane emits high-energy neutrinos, detected at 5.7σ significance — the first astrophysical source of such neutrinos to pass the 5σ discovery threshold.
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 →
IceCube detects high-energy neutrinos from the Milky Way's Galactic plane at 5.7σ post-trial significance, with the excess concentrated in the inner Galaxy.
T0 review reviewed 2026-08-01 challenge →
load-bearing objection Solid IceCube result; the 5.7σ Galactic-plane detection holds up, with the main caveats being the a posteriori inner-Galaxy excess and the usual lack of public event data.
High-energy neutrino emission from the Milky Way
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
Core claim
The paper's central claim is that the Galactic plane is a genuine emitter of TeV–PeV neutrinos: a predefined, global likelihood analysis of 1,069,454 events finds a signal at 5.7σ after accounting for the look-elsewhere effect, exceeding the 5σ threshold for discovery. The emission is extended along the Galactic plane and peaks in the inner Galaxy; the shower-dominated sample alone gives a local significance of 5.29σ, and adding starting and through-going tracks raises it to 5.97σ before trial correction. The authors show that the observed excess is not a fluctuation of the isotropic background, and that it is broadly consistent with diffuse emission predicted by models of cosmic-ray interac
What carries the argument
A joint unbinned maximum-likelihood fit that combines three event samples — shower-dominated, starting-track, and through-going-track — into one signal-plus-background model. The signal's spatial shape is fixed by four Galactic emission templates; the background is estimated from data by exploiting the detector's rotational symmetry, and significance is calibrated with hundreds of millions of right-ascension-randomized pseudo-experiments. The load-bearing technical improvement is a revised ice model — including birefringence, ice-layer undulations, and hole-ice angular response — which removes azimuthal artifacts and a zenith bias of up to about 10 degrees in shower reconstruction.
Load-bearing premise
The updated ice model and event reconstruction must have removed all detector-induced directional biases; if residual biases remain, reconstructed shower directions could be shifted toward the Galactic plane and mimic a signal.
What would settle it
Run the same 12-year shower analysis with an independent ice model that omits birefringence, or check the azimuthal distribution of reconstructed showers for residual alignment with the ice-flow direction; if the excess disappears, or a residual anisotropy along the ice-flow axis persists, the 5.7σ claim fails.
If this is right
- The Milky Way becomes the first known astrophysical source of high-energy neutrinos to pass the 5σ discovery threshold.
- The inner-Galaxy excess of roughly 60 events provides a new, neutrino-based handle on cosmic-ray transport and the distribution of interstellar gas.
- The fitted flux in the inner Galaxy is comparable, per solid angle, to the all-sky astrophysical neutrino intensity and is about 10% of the total astrophysical flux at ~20 TeV, making the Galactic component a measurable part of the sky.
- The data prefer templates like Fermi-LAT π0 and CRINGE over KRAγ; the π0 template requires a five-fold normalization boost, indicating that nominal models under-predict the observed flux.
- With more exposure and better angular resolution, future data can separate truly diffuse emission from unresolved source populations and test neutrino properties over kiloparsec baselines.
Where Pith is reading between the lines
- Editorial: the near-unity normalization of the CRINGE template, which includes an unresolved-source component, hints that a substantial part of the signal could come from point-like or compact Galactic sources rather than purely interstellar emission — a prediction that future point-source searches can test.
- Editorial: the same improved ice model used here could shift the significance of other IceCube results, such as point-source or diffuse-flux measurements, so re-analyses of older data might change slightly.
- Editorial: if the inner-Galaxy flux is confirmed, comparing it with gamma-ray maps of the same region could identify the specific cosmic-ray accelerators responsible, effectively using neutrinos as a beam dump for the Galaxy's PeVatrons.
- Editorial: the flavour composition of this Galactic signal, measured over kiloparsec baselines, could eventually constrain exotic neutrino physics such as decay or secret interactions, complementing shorter-baseline terrestrial experiments.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper reports an all-flavour IceCube analysis of 12 years of data, combining shower-dominated, starting-track, and through-going-track samples in a unified likelihood. The signal is modeled by four published Galactic emission templates (Fermi-LAT π0, KRAγ5, KRAγ50, CRINGE), with only the overall normalization fitted. For the Fermi-LAT π0 template the analysis finds a local significance of 5.97σ, corresponding to a global post-trial significance of 5.7σ after accounting for the look-elsewhere effect over the four templates. The excess is concentrated toward the inner Galaxy; a separate counting illustration finds 217 shower events with Ê>5 TeV and |b̂|<15° against a background expectation of 154.4±4.1. The paper also presents post-fit inner-Galaxy spectra, normalization factors for each template, and cross-checks with the previous 10-year analysis using the older reconstruction.
Significance. If correct, this is the first 5σ detection of a specific astrophysical source class in high-energy neutrinos, establishing the Milky Way as a resolved neutrino source and opening a new window on Galactic cosmic-ray transport. The analysis has important strengths: the main likelihood is predefined, the background is data-driven and right-ascension randomized, the global p-value is based on 215 million pseudo-experiments, the four templates are external published models with only a single fitted normalization, and the templates and best-fit normalizations are publicly released. I specifically considered the concern that residual ice-model directional bias could mimic a Galactic-plane signal. That concern does not land: at the South Pole, any local-azimuth-dependent acceptance or reconstruction bias averages out in right ascension for time-integrated observations, because every celestial source is observed over the full cycle of local azimuths; moreover, the cross-check with the previous reconstruction (Supplement Table 3) yields a comparable shower-only significance (5.42σ), making it implausible that the signal is an artifact of the updated ice model. A rotated-template test would b
minor comments (4)
- [Abstract] The abstract's '217 shower events... compared with an expected background of 154.4±4.1' belongs to a counting analysis whose on/off thresholds were optimized a posteriori (Methods, 'Optimisation of the data residual in Galactic longitude'). The paper discloses this, but the abstract should explicitly state that this is a post-hoc illustrative study, not a second predefined measurement, to avoid readers interpreting it as an independent discovery with its own significance.
- [Methods, Likelihood ratio construction] The global 5.7σ significance is obtained by fitting the 215M-pseudo-experiment background TS distribution with a chi-square distribution with k=1.16 and then reading the tail. The observed TS lies beyond the most extreme pseudo-experiment (1/215M corresponds to about 5.65σ), so the reported p-value relies on a small extrapolation. Please report the empirical p-value as an upper limit and quantify the systematic uncertainty of the tail fit, for example by comparing alternative tail models or fitting over different TS ranges.
- [Methods, Likelihood ratio construction] Text reads 'Earth's rotation changes only changes only the local azimuthal angle of a celestial object' — duplicated phrase 'changes only'. Please fix this typo.
- [Supplement Table 1 / main text] The ESTES sample is listed with 'statistical uncertainties only' because its systematic simulation set does not exist. This is a limitation that also affects the combined-fit uncertainty and should be stated explicitly in the main text where the combined normalization and significance are discussed, not only in the supplement table footnote.
Circularity Check
Central 5.7σ detection is not circular: templates are external, only the normalisation is fitted, and the null is data-driven via RA-randomisation with look-elsewhere correction. One disclosed a-posteriori-optimised residual is flagged as partial, peripheral reduction-by-construction.
specific steps
-
fitted input called prediction
[Main text, Fig. 3 paragraph; Methods, 'Optimisation of the data residual in Galactic longitude']
"The Galactic templates, with best-fit normalizations applied, predict 30–40 excess events in the same bin. ... The binning and event-selection criteria used for the Galactic longitude profiles in Fig. 3 and Extended Data Fig. 9 were determined a posteriori, based on the best-fit Fermi–LAT π0 template obtained in the likelihood analysis."
The quoted 30–40 'predicted' excess events are the fitted template scaled by the best-fit n_s obtained from the same events, and the thresholds (bmax = 15°, Emin = 5 TeV) underlying the abstract's '217 shower events ... versus 154.4 ± 4.1' were chosen by maximising the FoM in Eq. 11 with signal injected at that same data-best-fit normalisation. The residual 62.6 ± 15.3 and the overlaid 'expectations' are therefore partly constructed from the fit rather than independent predictions. The paper explicitly discloses this a-posteriori optimisation and states it does not enter the likelihood fit, so the main 5.7σ result is not affected; this is peripheral, partial circularity.
full rationale
Walking the derivation chain: the four emission templates (Fermi-LAT π0, KRAγ5, KRAγ50, CRINGE) are externally published models anchored to gamma-ray and cosmic-ray data; the likelihood (Eqs. 9–10) fits only the overall normalisation n_s, and the signal-PDF shapes are fixed by the templates and MC. The background is data-driven, exploiting South-Pole azimuthal symmetry, and the null is built by RA-randomisation; the look-elsewhere effect over the four templates is applied with ~215 million pseudo-experiments (Extended Data Fig. 4). Thus the 5.7σ post-trial claim is a spatial-correlation test whose inputs are not the output. The self-citations for improved ice modelling and reconstruction (refs 22,24,25,26) are load-bearing in the sense that they explain the improved shower significance, but they are supported by external LED calibration data, by the direct data check of azimuthal uniformity (Extended Data Fig. 2), and crucially by an independent cross-check: Supplement Table 3 reproduces >5σ (5.42σ for showers) using the previous, uncorrected reconstruction, so the detection does not reduce to the self-cited model chain. CRINGE has an IceCube-affiliated co-author (Wiebusch), but all four templates yield >5σ local significance, so no conclusion rests solely on that template. The only exhibited reduction-by-construction is the Fig. 3 / abstract residual, whose cuts and 'predictions' are a-posteriori tuned and scaled from the same fit; this is fully disclosed in Methods and is peripheral to the central 5.7σ claim. Accordingly, the circularity burden is low: the central derivation is self-contained against external benchmarks, with one minor disclosed item and one non-load-bearing author-overlap, yielding a score of 2.
Axiom & Free-Parameter Ledger
free parameters (2)
- Signal normalization (ns / template flux scale) =
Fermi π0 ×4.99; KRAγ5 ×0.53; KRAγ50 ×0.37; CRINGE ×0.91
- Longitude-profile selection thresholds (bmax, Emin) =
Showers bmax=15°, Emin=5 TeV; tracks bmax=8.5°, Emin=0.1 TeV
axioms (5)
- domain assumption Neutrino flavour ratio at Earth is νe:νμ:ντ = 1:1:1
- domain assumption Background is independent of right ascension
- domain assumption The four Galactic emission templates approximate the true neutrino morphology
- ad hoc to paper The updated ice model (birefringence, layer undulations, hole-ice) is correct
- domain assumption Detector simulation and effective areas are accurate
Cite this review
Pith. "Pith review of High-energy neutrino emission from the Milky Way." pith.science (2026). https://pith.science/paper/6JHJJFRK
@misc{pith2026260725966,
author = {Pith},
title = {Pith review of: High-energy neutrino emission from the Milky Way},
year = {2026},
howpublished = {\url{https://pith.science/paper/6JHJJFRK}},
note = {Machine review of arXiv:2607.25966}
}
read the original abstract
The Milky Way hosts astrophysical objects that accelerate cosmic rays to energies beyond the reach of terrestrial particle accelerators. It remains a longstanding goal to locate the sites of these powerful Galactic engines and understand how cosmic rays propagate through the Galaxy, leading to the production of high-energy neutrinos. In this paper, we combine event morphologies characteristic of all three neutrino flavours and apply recent improvements in ice modelling, calibration and reconstruction to 12 years of IceCube data. With a predefined, global analysis we establish high-energy neutrino emission from the Galactic plane at 5.7 $\sigma$ significance. A further study shows that the inner region of the Galaxy is a prominent neutrino source, with 217 shower events with visible energy above 5 TeV compared with an expected background of 154.4 $\pm$ 4.1. These results herald a new era of Galactic multi-messenger astronomy, creating new opportunities to study cosmic-ray propagation and probe neutrino properties over kiloparsec distances.
Figures
Reference graph
Works this paper leans on
-
[1]
J.750,3 (2012)
Ackermann, M.et al.Fermi-LAT Observations of the Diffuse Gamma-Ray Emission: Impli- cations for Cosmic Rays and the Interstellar Medium.Astrophys. J.750,3 (2012)
2012
-
[2]
& Valli, M
Gaggero, D., Grasso, D., Marinelli, A., Urbano, A. & Valli, M. The gamma-ray and neutrino sky: A consistent picture of Fermi-LAT, Milagro, and IceCube results.Astrophys. J. Lett.815, L25 (2015)
2015
-
[3]
& Wiebusch, C
Schwefer, G., Mertsch, P. & Wiebusch, C. Diffuse Emission of Galactic High-energy Neutri- nos from a Global Fit of Cosmic Rays.Astrophys. J.949,16 (2023)
2023
-
[4]
Abbasi, R.et al.Observation of high-energy neutrinos from the Galactic plane.Science380, 1338–1343 (2023)
2023
-
[5]
Abbasi, R.et al.In situ estimation of ice crystal properties at the South Pole using LED calibration data from the IceCube Neutrino Observatory.The Cryosphere18,75–102 (2024)
2024
-
[6]
Abbasi, R.et al.An improved mapping of ice layer undulations for the IceCube Neutrino Observatory. inProc. 38th Int Cosmic Ray Conf. (ICRC2023)444(2023), 975
2023
-
[7]
Abbasi, R.et al.Improved modeling of in-ice particle showers for IceCube event reconstruc- tion.JINST19,P06026 (2024)
2024
-
[8]
G.et al.Constraints on Galactic Neutrino Emission with Seven Years of IceCube Data.Astrophys
Aartsen, M. G.et al.Constraints on Galactic Neutrino Emission with Seven Years of IceCube Data.Astrophys. J.849,67 (2017)
2017
-
[9]
G.et al.Search for Sources of Astrophysical Neutrinos Using Seven Years of IceCube Cascade Events.Astrophys
Aartsen, M. G.et al.Search for Sources of Astrophysical Neutrinos Using Seven Years of IceCube Cascade Events.Astrophys. J.886,12 (2019). 10
2019
-
[10]
J.998,37 (2026)
Abbasi, R.et al.Time-integrated Southern-sky Neutrino Source Searches with 10 yr of Ice- Cube Starting-track Events at Energies Down to 1 TeV.Astrophys. J.998,37 (2026)
2026
-
[11]
M.et al.Galactic and Extragalactic Analysis of the Astrophysical Muon Neutrino Flux with 12.3 years of IceCube Track Data
Fuerst, P. M.et al.Galactic and Extragalactic Analysis of the Astrophysical Muon Neutrino Flux with 12.3 years of IceCube Track Data. inProc. 38th Int Cosmic Ray Conf. (ICRC2023) 444(2023), 1046
2023
-
[12]
Neste, L.et al.Measuring the Neutrino Flux in Segments along the Galactic Plane with IceCube.PoSICRC2025,1130 (2025). 11
2025
This paper was first reviewed by deepseek-v4-flash on August 1, 2026.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.