arxiv: 2604.00111 · v2 · submitted 2026-03-31 · 🌌 astro-ph.CO · astro-ph.GA· astro-ph.IM· astro-ph.SR· gr-qc

Recognition: unknown

Eppur non si trovano Vol. 2: No Planetary-mass Primordial Black Holes toward the Andromeda Galaxy

Przemek Mr\'oz , Andrzej Udalski

Authors on Pith no claims yet

Pith reviewed 2026-05-08 02:24 UTC · model gemini-3-flash-preview

classification 🌌 astro-ph.CO astro-ph.GAastro-ph.IMastro-ph.SRgr-qc PACS 95.35.+d98.35.Gi98.62.Sb

keywords Primordial black holesDark matterGravitational microlensingAndromeda GalaxyVariable starsDifference Image Analysis

0 comments

The pith

The claimed discovery of planetary-mass primordial black holes in the Andromeda Galaxy is a misidentification of variable stars.

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

A recent analysis of high-cadence imagery of the Andromeda Galaxy suggested that twelve short-duration flashes were gravitational microlensing events caused by primordial black holes. If true, this would have implied that these black holes account for the universe's dark matter. This paper re-examines the same data using an independent image subtraction technique and finds that every candidate is actually a known type of variable star. By demonstrating that these objects show recurring or asymmetric patterns, the authors argue that there is no evidence for a significant population of planetary-mass black holes in the region.

Core claim

The authors show that the twelve microlensing candidates identified in previous work are misclassified variable stars, primarily RR Lyrae stars and eclipsing binaries. By applying an independent difference image analysis pipeline to the original data, they reveal that the light curves of these objects exhibit asymmetries and multi-night variability that are incompatible with the physics of gravitational microlensing. This finding removes the primary evidence for a large population of planetary-mass primordial black holes acting as dark matter in the local group.

What carries the argument

Difference Image Analysis (DIA). This is a method that subtracts a reference image of a star field from a new observation to isolate changes in brightness. It is used here to distinguish the unique, symmetric signature of a one-time microlensing event from the repeating or asymmetric patterns of variable stars in the crowded environment of a distant galaxy.

If this is right

Constraints on the abundance of primordial black holes in the mass range of 10^-6 to 10^-4 solar masses remain at their previous low levels.
The search for dark matter must continue to focus on other candidates, as planetary-mass black holes are ruled out as a primary component.
Future microlensing surveys must incorporate multi-night baselines to effectively filter out stellar variables that mimic short-duration transients.

Where Pith is reading between the lines

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

The misidentification in the original study likely occurred because the observation window was too narrow to catch the full cycle of the variable stars' periodic pulses.
The result suggests that high-cadence sampling can create 'aliasing' where periodic signals are mistaken for single peaks if the temporal baseline is insufficient.
This reinforces the model that the dark matter halo is not composed of compact objects in the planetary mass range, pushing researchers back toward particle-based or extremely high/low mass theories.

Load-bearing premise

The independent image processing pipeline correctly handles the extremely dense star fields of the Andromeda Galaxy without introducing its own systematic errors.

What would settle it

If long-term monitoring of these twelve specific locations shows that the brightness changes never repeat and maintain a perfectly symmetric profile, the identification as variable stars would be invalidated.

Figures

Figures reproduced from arXiv: 2604.00111 by Andrzej Udalski, Przemek Mr\'oz.

**Figure 1.** Figure 1: Full Subaru HSC light curve of the candidate microlensing event iden view at source ↗

**Figure 2.** Figure 2: Light curve of the candidate microlensing event identified by Niikura view at source ↗

**Figure 3.** Figure 3: Finding charts for the microlensing event candidates found by Sugiyama view at source ↗

**Figure 4.** Figure 4: Stacks of ≈ 15 difference images centered on the peaks of the purported microlensing events. The position of each variable is marked by the white tick marks. Images are 30′′ ×30′′ , North is up and East is to the left view at source ↗

**Figure 5.** Figure 5: Full Subaru HSC light curve of the candidate microlensing event #01 view at source ↗

**Figure 6.** Figure 6: Full Subaru HSC light curve of the candidate microlensing event #02 view at source ↗

**Figure 7.** Figure 7: Full Subaru HSC light curve of the candidate microlensing event #03 view at source ↗

**Figure 8.** Figure 8: Full Subaru HSC light curve of the candidate microlensing event #04 view at source ↗

**Figure 9.** Figure 9: Full Subaru HSC light curve of the candidate microlensing event #05 view at source ↗

**Figure 10.** Figure 10: Full Subaru HSC light curve of the candidate microlensing event #06 view at source ↗

**Figure 11.** Figure 11: Full Subaru HSC light curve of the candidate microlensing event #07 view at source ↗

**Figure 12.** Figure 12: Full Subaru HSC light curve of the candidate microlensing event #08 view at source ↗

**Figure 13.** Figure 13: Full Subaru HSC light curve of the candidate microlensing event #09 view at source ↗

**Figure 14.** Figure 14: Full Subaru HSC light curve of the candidate microlensing event #10 view at source ↗

**Figure 15.** Figure 15: Full Subaru HSC light curve of the candidate microlensing event #11 view at source ↗

**Figure 16.** Figure 16: Full Subaru HSC light curve of the candidate microlensing event #12 view at source ↗

**Figure 17.** Figure 17: Phase-folded light curves for ten candidates from Sugiyama view at source ↗

**Figure 18.** Figure 18: The most stringent 95% upper limits from the gravitational microlensing view at source ↗

read the original abstract

A recent preprint by Sugiyama et al. reported the discovery of twelve candidates for short-timescale (less than one day) gravitational microlensing events based on high-cadence photometric observations of the Andromeda Galaxy (M31) using the Subaru Hyper Suprime-Cam. These detections were attributed to a large population of planetary-mass primordial black holes (PBHs) that could account for the entirety of the dark matter in the Milky Way and M31 halos. However, these results are in clear tension with previous searches for short-timescale microlensing events toward the Magellanic Clouds, such as those by the OGLE survey. In addition, both the temporal and spatial distributions of the Subaru candidates are inconsistent with expectations for microlensing events. Here, we reanalyze the Subaru data using an independent difference image analysis photometric pipeline. We find that all twelve candidates identified by Sugiyama et al. exhibit asymmetric light curves and/or variability on multiple nights of Subaru observations. Our analysis reveals that among them ten objects are RR Lyrae stars, one is an eclipsing binary, and one is an unclassified variable star. We find no compelling evidence for short-timescale microlensing events among the candidates identified in the Subaru data set, nor for a significant population of planetary-mass PBHs as dark matter components. Our results underscore the necessity of robust variable-star rejection in high-cadence microlensing searches using large telescopes.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The claim that planetary-mass primordial black holes explain dark matter is essentially refuted; the candidates were misidentified variable stars.

read the letter

You might have seen the Sugiyama et al. paper from a few months back claiming they found 12 planetary-mass primordial black holes toward Andromeda. It made a splash because it suggested PBHs could account for all dark matter. This paper by Mróz and Udalski effectively closes that door.

They re-analyzed the same Subaru dataset using the OGLE pipeline. The original team looked for transients in a very narrow window and missed the fact that these objects were actually periodic. Mróz and Udalski found that all 12 candidates show up on other nights. Specifically, ten of them are textbook RR Lyrae stars, one is an eclipsing binary, and one is just a generic variable star.

The evidence is as solid as it gets in photometry. When you look at the phased light curves in the paper, the sawtooth shape of the RR Lyrae is unmistakable. Plus, the magnitudes (around r = 25.4) are exactly where you would expect RR Lyrae to be in the Andromeda Galaxy. Microlensing is a one-time event; if a source repeats or shows periodicity, it is not a lens. These repeat.

There are not any real soft spots here in terms of the logic. It is a straightforward technical debunking. They used a different image subtraction method (DIA) which seems to have handled the crowded M31 fields better, or at least allowed them to track the stars over a longer baseline than the original team did. The stress-test mention of PSF-matching residuals causing spurious periods is a fair thing to check, but the light curves here are too clean and too physically consistent with RR Lyrae (both in shape and absolute magnitude) to be artifacts.

This is a must-read if you are tracking the dark matter debate. It is a cautionary tale about high-cadence surveys and the necessity of checking for variable star contamination before making cosmological claims. It is a definitive correction to the record and deserves a serious referee.

Referee Report

2 major / 3 minor

Summary. This manuscript re-evaluates the 12 planetary-mass primordial black hole (PBH) microlensing candidates recently reported by Sugiyama et al. (2024) using Subaru/HSC data. By employing an independent difference image analysis (DIA) pipeline based on the OGLE software and incorporating data from multiple observation nights (2016–2020), the authors demonstrate that all 12 candidates are actually known or newly identified variable stars. Specifically, they identify 10 RR Lyrae stars, one eclipsing binary, and one unclassified variable. The multi-night variability and characteristic light-curve shapes (phased with periods between 0.29 and 0.77 days) are physically incompatible with the one-time, symmetric transient nature of gravitational microlensing, thereby nullifying the claim of a significant planetary-mass PBH population based on this dataset.

Significance. The paper provides a crucial corrective to a high-impact claim in the field of dark matter. The identification of primordial black holes as the primary constituent of dark matter would be a fundamental discovery; however, this manuscript shows that the Sugiyama et al. results were likely contaminated by stellar variability. The significance lies in the methodological rigor: using an independent pipeline (OGLE DIA), multi-year archival data to check for periodicity, and physical consistency checks (standard candle magnitudes for RR Lyrae at M31 distance). This is a textbook example of how high-cadence surveys require multi-epoch follow-up or archival cross-matching to distinguish microlensing from the vast background of intrinsic stellar variables.

major comments (2)

[Section 3.1, Table 1] The authors derive precise periods (e.g., 0.51125 d for Object 1) using data spanning multiple years. To ensure the robustness of the variable star classification, the manuscript should explicitly state the period-searching algorithm employed (e.g., Lomb-Scargle, Analysis of Variance) and provide a measure of the statistical significance or 'alias risk' for these periods, given the large gaps between Subaru runs (S16A to S20B).
[Section 3.2, Paragraph 1] The claim that the observed magnitudes ($r \approx 25.4$) are consistent with RR Lyrae stars at M31 distance is central to the debunking. The authors should provide a brief calculation or reference comparing this observed magnitude to the expected $M_r$ of RR Lyrae stars, adjusted for the M31 distance modulus ($μ \approx 24.4$) and typical extinction, to strengthen the 'standard candle' argument for this identification.

minor comments (3)

[Figure 1 and 2] The phased light curves are highly convincing. However, it would be helpful to include the original (unphased) data for at least one representative object in an appendix or a small inset to illustrate the sparsity of the sampling relative to the multi-year baseline.
[Abstract & Introduction] The manuscript mentions that the 'spatial distributions' of the Subaru candidates were inconsistent with microlensing expectations. While the periodicity argument is the 'smoking gun,' the paper would benefit from a small plot showing the spatial distribution of these 12 stars relative to the M31 disk/halo structure to support this secondary claim.
[Section 3.1, Object 12] Object 12 is listed as 'unclassified variable.' While its variability across multiple nights (Figure 2) is sufficient to rule out microlensing, a brief mention of why it does not fit the RR Lyrae or EB categories (e.g., irregular timing or unusual color) would be informative.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their positive assessment of our manuscript and for recognizing the methodological rigor of our re-analysis. The referee’s comments regarding the period-finding procedure and the standard candle nature of the candidates are well-taken. We have incorporated specific details and calculations to address these points, which we believe strengthen the physical arguments presented in the paper.

read point-by-point responses

Referee: [Section 3.1, Table 1] The authors derive precise periods (e.g., 0.51125 d for Object 1) using data spanning multiple years. To ensure the robustness of the variable star classification, the manuscript should explicitly state the period-searching algorithm employed (e.g., Lomb-Scargle, Analysis of Variance) and provide a measure of the statistical significance or 'alias risk' for these periods, given the large gaps between Subaru runs (S16A to S20B).

Authors: We have updated Section 3.1 to specify that we used the Analysis of Variance (AoV; Schwarzenberg-Czerny 1989) algorithm for period searching, which is particularly robust for non-sinusoidal light curves like those of RR Lyrae stars. Regarding the 'alias risk': while the multi-year gaps between Subaru runs (S16A, S19A, S20B) are indeed large, the risk of daily aliasing is significantly mitigated by the high cadence within individual nights. For example, the S16A run provides nearly 7 hours of continuous monitoring. The observed 'slope' or rate of change during these windows acts as a strong constraint on the period, as the phase must evolve consistently with the intra-night data. This intra-night 'lever arm' allows us to distinguish between aliases that would otherwise be degenerate in sparser datasets. We have added a sentence to the manuscript clarifying how the intra-night cadence assists in period disambiguation. revision: yes
Referee: [Section 3.2, Paragraph 1] The claim that the observed magnitudes ($r \approx 25.4$) are consistent with RR Lyrae stars at M31 distance is central to the debunking. The authors should provide a brief calculation or reference comparing this observed magnitude to the expected $M_r$ of RR Lyrae stars, adjusted for the M31 distance modulus ($μ \approx 24.4$) and typical extinction, to strengthen the 'standard candle' argument for this identification.

Authors: We agree that an explicit calculation strengthens the identification. Taking the distance modulus to M31 as $\mu \approx 24.44$ (e.g., Ribas et al. 2005) and the typical absolute magnitude of RR Lyrae stars as $M_r \approx +0.5$ to $+0.6$ (depending on metallicity), the expected apparent magnitude is $r \approx 25.0$. Accounting for a typical foreground Galactic extinction toward M31 of $A_r \approx 0.17$ (Schlafly & Finkbeiner 2011) and potential moderate extinction within M31 itself, the expected observed magnitude is $r \approx 25.2-25.5$. Our identified candidates, centered around $r \approx 25.4$, are in excellent agreement with this range. We have added this arithmetic and the relevant references to Section 3.2 to make this consistency more transparent to the reader. revision: yes

Circularity Check

0 steps flagged

No circularity: independent empirical re-analysis of observational data.

full rationale

The paper presents an independent re-reduction of the Subaru Hyper Suprime-Cam dataset using the OGLE Difference Image Analysis (DIA) pipeline. The core finding—that twelve previously reported microlensing candidates are in fact variable stars (primarily RR Lyrae and eclipsing binaries)—is reached through an empirical identification of multi-night periodicity and light-curve asymmetry in the processed data. The analysis does not rely on self-defining assumptions or load-bearing self-citations to force its conclusions; rather, it identifies physical characteristics in the light curves (such as specific periods and magnitudes consistent with the distance to M31) that are incompatible with the microlensing hypothesis. The methodology is a standard observational cross-check where the 'prediction' of variable star nature is a falsifiable discovery made through data analysis, not a result guaranteed by the pipeline's construction or author-specific theorems.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The paper uses standard astronomical principles to identify variables. It does not introduce new physical constants or entities; it identifies existing ones incorrectly labeled by others.

axioms (2)

domain assumption Gravitational microlensing events are singular, non-repeating events.
Used in Section 1 to define the criteria for rejecting PBH candidates if they show multi-night variability.
domain assumption Light curve morphology (shape and period) can definitively classify RR Lyrae stars.
Invoked in Section 3 to reclassify the 12 candidates as variable stars.

pith-pipeline@v0.9.0 · 7269 in / 1510 out tokens · 41948 ms · 2026-05-08T02:24:38.959679+00:00 · methodology

discussion (0)

Forward citations

Cited by 3 Pith papers

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

Microlensing of fast and slow compact objects
astro-ph.CO 2026-04 unverdicted novelty 7.0

Microlensing surveys constrain fast and slow compact objects at masses and densities differing by orders of magnitude from dark matter limits due to speed-mass degeneracy in Einstein crossing times.
Primordial Black Hole from Tensor-induced Density Fluctuation: First-order Phase Transitions and Domain Walls
astro-ph.CO 2026-05 unverdicted novelty 6.0

Tensor perturbations from first-order phase transitions and domain wall annihilation induce curvature fluctuations at second order that form primordial black holes, allowing asteroid-mass PBHs to comprise all dark mat...
In-depth analysis of the clustering of dark matter particles around primordial black holes. Part III: CMB constraints
astro-ph.CO 2026-04 unverdicted novelty 4.0

CMB data limits the s-wave annihilation cross section of thermal dark matter particles to ≲ 10^{-30} cm³/s scaled by PBH fraction and mass for PBHs heavier than ~10^{-10} solar masses.

Reference graph

Works this paper leans on

18 extracted references · 3 canonical work pages · cited by 3 Pith papers

[1]

Alcock, C.,et al.1996,ApJ,471,

Aihara, H.,et al.2018,PASJ,70, S4. Alcock, C.,et al.1996,ApJ,471,

2018
[2]

Baba, H.,et al.2002, in Astronomical Data Analysis Software and Systems XI,Astronomical Society of the Pacific Conference Series, vol

Alcock, C.,et al.1998,ApJL,499, L9. Baba, H.,et al.2002, in Astronomical Data Analysis Software and Systems XI,Astronomical Society of the Pacific Conference Series, vol. 281, eds. D.A. Bohlender, D. Durand, and T.H. Handley,

1998
[3]

Chambers, K.C.,et al.2016,arXiv e-prints, arXiv:1612.05560

Blas, D.,et al.2026,arXiv e-prints, arXiv:2602.12252. Chambers, K.C.,et al.2016,arXiv e-prints, arXiv:1612.05560. Craig, M.,et al.2025, astropy/ccdproc: 2.5.1: The final 2.x release. Crotts, A.P.S. 1992,ApJL,399, L43. Domènech, G., Pi, S., and Wang, A. 2026,arXiv e-prints, arXiv:2602.24061. Furusawa, H.,et al.2018,PASJ,70, S3. Gaia Collaboration,et al.201...

work page arXiv 2026
[4]

1994,A&A,286,

Jetzer, P. 1994,A&A,286,

1994
[5]

2026,arXiv e-prints, arXiv:2602.09558

Kasai, K., Kawasaki, M., Murai, K., and Neda, S. 2026,arXiv e-prints, arXiv:2602.09558. Kawanomoto, S.,et al.2018,PASJ,70,

work page arXiv 2026
[6]

Li, S., Riess, A.G., Busch, M.P., Casertano, S., Macri, L.M., and Yuan, W

Komiyama, Y .,et al.2018,PASJ,70, S2. Li, S., Riess, A.G., Busch, M.P., Casertano, S., Macri, L.M., and Yuan, W. 2021,ApJ,920,

2018
[7]

Mróz, P.,et al.2017,Nature,548,

Miyazaki, S.,et al.2018,PASJ,70, S1. Mróz, P.,et al.2017,Nature,548,

2018
[8]

Niikura, H., Takada, M., Yokoyama, S., Sumi, T., and Masaki, S

Narloch, W.,et al.2024,A&A,689, A138. Niikura, H., Takada, M., Yokoyama, S., Sumi, T., and Masaki, S. 2019a,Phys. Rev. D,99, 083503. Niikura, H.,et al.2019b,Nature Astronomy,3,

2024
[9]

Renault, C.,et al.1998,A&A,329,

Renault, C.,et al.1997,A&A,324, L69. Renault, C.,et al.1998,A&A,329,

1997
[10]

1993,PASP,105,

Schechter, P.L., Mateo, M., and Saha, A. 1993,PASP,105,

1993
[11]

1989,MNRAS,241,

Schwarzenberg-Czerny, A. 1989,MNRAS,241,

1989
[12]

Soszy´nski, I.,et al.2014,Acta Astron.,64,

2014
[13]

Soszy´nski, I.,et al.2016,Acta Astron.,66,

2016
[14]

Sugiyama, M

Sugiyama, S., Takada, M., Yasuda, N., and Tominaga, N. 2026,arXiv e-prints, arXiv:2602.05840. Sumi, T.,et al.2023,AJ,166,

work page arXiv 2026
[15]

2003,Acta Astron.,53,

Udalski, A. 2003,Acta Astron.,53,

2003
[16]

2015,Acta Astron.,65,

Udalski, A., Szyma´nski, M.K., and Szyma´nski, G. 2015,Acta Astron.,65,

2015
[17]

2000,Acta Astron.,50,

Wo´ zniak, P.R. 2000,Acta Astron.,50,

2000
[18]

V ol. 013 6985.8 6985.9 23.0 23.5 24.0 24.5 25.0Magnitude 2014-11-24 8016.9 8017.0 2017-09-20 9143.8 9143.9 23.0 23.5 24.0 24.5 25.0Magnitude 2020-10-21 9144.8 9144.9 2020-10-22 9164.8 9164.9 23.0 23.5 24.0 24.5 25.0Magnitude 2020-11-11 9165.8 9165.9 2020-11-12 9167.8 9167.9 23.0 23.5 24.0 24.5 25.0Magnitude 2020-11-14 9173.8 9173.9 2020-11-20 Candidate (...

2014