pith. sign in

arxiv: 2602.00301 · v2 · submitted 2026-01-30 · 🌌 astro-ph.CO

Studying dark gaps in Ly-α forest transmission with large reionization simulations

Barun Maity , Frederick B. Davies , Benedetta Spina , Sarah E. I. Bosman This is my paper

Pith reviewed 2026-05-16 08:52 UTC · model grok-4.3

classification 🌌 astro-ph.CO
keywords Ly-alpha forestdark gapscosmic reionizationintergalactic mediumcosmological simulationsquasar spectraionizing photon mean free pathultraviolet background
0
0 comments X

The pith

Simulations show that reionization ending near redshift 5.4 best matches the observed distribution of long dark gaps in the Ly-alpha forest.

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

The paper uses large-volume cosmological simulations to generate synthetic quasar spectra and test how the timing of reionization, the temperature of the intergalactic gas, and the distance ionizing photons travel affect the sizes and numbers of dark gaps where Ly-alpha transmission drops below 5 percent. By comparing the cumulative distribution of these gaps across redshift bins to real spectra from the XQR-30 survey, the authors find that models finishing reionization slightly later produce gap statistics closest to the data. A model with a short, constant mean free path for ionizing photons is inconsistent with the observations at lower redshifts. The results give qualitative support for reionization completing at or below redshift 5.7, although they leave room for an earlier end accompanied by stronger fluctuations in the ultraviolet background. The same models still fail to reproduce the unusually large-scale flux correlations reported in the data.

Core claim

Using an efficient ~1 Gpc simulation framework, the authors construct model variants that vary reionization completion redshift, IGM temperature, and ionizing-photon mean free path, then extract synthetic sightlines and compare the predicted cumulative distribution of dark gaps (transmitted flux below 0.05) to observed spectra in Delta-z = 0.2 intervals. The z approximately 5.4 completion scenario matches the data most closely while a short constant mean-free-path model is disfavored at lower redshifts; the distribution of long gaps (L greater than or equal to 30 h inverse cMpc) yields the same qualitative conclusion. These outcomes support an extended reionization end at z less than or 5.7,

What carries the argument

The cumulative distribution of dark-gap lengths in synthetic Ly-alpha forest sightlines, generated by varying reionization redshift, IGM temperature, and mean free path in large-volume simulations.

If this is right

  • A reionization completion redshift near 5.4 reproduces the observed dark-gap statistics more closely than earlier or later endings.
  • Models with a short constant ionizing-photon mean free path are inconsistent with the data at redshifts below 5.6.
  • The results qualitatively favor scenarios in which reionization finishes at or below redshift 5.7.
  • An earlier reionization completion combined with stronger post-reionization ultraviolet background fluctuations cannot yet be ruled out.
  • No model variant reproduces the strong large-scale Ly-alpha flux correlations seen in the observations.

Where Pith is reading between the lines

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

  • Dark-gap statistics at z less than 5.6 could become a useful independent probe for distinguishing between late reionization and fluctuating ultraviolet background models once larger samples are available.
  • Including explicit inhomogeneous reionization morphology in future simulation suites may be required to simultaneously match both gap sizes and the reported large-scale correlations.
  • If the large-scale correlations persist in new data, they would point to missing physics such as rare bright sources or large-scale temperature fluctuations not captured by the current parameter variations.

Load-bearing premise

That the main drivers of dark-gap statistics are captured by changing only the global reionization redshift, gas temperature, and photon mean free path without needing patchy reionization geometry or extra sources of ultraviolet fluctuations.

What would settle it

New quasar spectra that show a dark-gap size distribution at z approximately 5.2-5.4 clearly inconsistent with the predictions of the z approximately 5.4 reionization model while still matching the short mean-free-path case.

Figures

Figures reproduced from arXiv: 2602.00301 by Barun Maity, Benedetta Spina, Frederick B. Davies, Sarah E. I. Bosman.

Figure 1
Figure 1. Figure 1: Different observables for the various model scenarios assumed in this study. From left to right, the panels show redshift evolution of global neutral fraction (Q V HI), mean IGM temperature (T0), index of temperature-density relation (γ) and the effective photon mean free path (λ0). We also show various constraints on these quantities, as suggested by recent studies, i.e, constraints on neutral fraction (D… view at source ↗
Figure 2
Figure 2. Figure 2: Lightcone snapshots for three different cases in (top: fiducial, middle: cons mfp / wo neutral, and bottom: low temp) in three rows. The columns correspond to density (∆), neutral fractions (xHI), UVB fluctuations (ΓHI/⟨ΓHI⟩), temperature (T), and flux (F). The colorbars have been shown in logarithmic scales. The rest of the scenarios has been shown in Appendix A. eter (λ0) evolves from 4 h −1 cMpc at z = … view at source ↗
Figure 3
Figure 3. Figure 3: The cumulative probability distribution functions (CDFs) of dark gap lengths at different redshift ranges (within an interval of ∆z = 0.2), corresponding to the set of different model variants, discussed in section 2.2. The solid lines denote the derived distribution from the observed data, while the dashed lines are the corresponding predictions using model skewers. The shaded region signifies 68% uncerta… view at source ↗
Figure 4
Figure 4. Figure 4: Fraction of skewers with dark gap length, L ≥ 30 h −1 cMpc (F30) as a function of redshift (z) for the different model variants, discussed in section 2.2. The red lines are derived distributions from the observed spectra. The blue dashed lines denote the mean distribution after averaging over skewer realizations from the model suites. The shaded regions show the corresponding 68% and 95% uncertainties. In … view at source ↗
Figure 5
Figure 5. Figure 5: Correlation coefficients of the transmission flux between redshift ranges considered in this study (z = 5.0 − 6.1). The panels show the different scenarios, as discussed in section 2.2. The correlation matrix derived from 67 quasar sigtlines (including E-XQR-30 samples) has been reported in Spina et al. (2025). patches. Otherwise, we mask those pixels from the analysis pro￾cedures. 4.2. Evolution of dark g… view at source ↗
Figure 6
Figure 6. Figure 6: Flux distribution as a function of redshift, for the different scenarios as discussed in section 2.2. The bands show the 68% and 95% intervals of the flux distributions in the extracted skewers. The mean fluxes of the coeval simulation boxes have been matched with the observed mean, shown in black points (Bosman et al. 2022). temperature model degrades the match as it provides a slightly earlier reionizati… view at source ↗
read the original abstract

The physical conditions of the intergalactic medium (IGM) during the final stages of cosmic reionization ($z\sim5.0-6.0$) are not yet fully understood. Recent reports of unexpectedly large-scale ($\ge 150 h^{-1}\mathrm{cMpc}$) correlation in Ly-$\alpha$ transmission flux using extended XQR-30 quasar spectra pose interesting challenges on the reionization end stages. In this work, we investigate the Ly-$\alpha$ forest dark-gap distribution (defined as regions with transmitted flux below 0.05) as another sensitive tracer of the IGM, using an efficient, large-volume ($\sim 1 ~\mathrm{Gpc}$) simulation framework. By constructing a suite of physically motivated model variants (i.e, varying the reionization redshift, IGM temperature, and ionizing-photon mean free path), we generate synthetic sightlines and compare their predicted cumulative distribution of dark gaps with that of observed spectra (at redshift intervals of $\Delta z=0.2$). We find that most of the models achieve qualitatively consistent agreement with the data. The scenario involving a slightly later reionization completion ($z\sim 5.4$) provides the closest match, while a short constant mean free path model is disfavored by the data at lower redshifts. These findings give qualitative support for the emerging scenario of reionization end extending to $z\le5.7$, although they can not rule out a slightly early reionization with enhanced post-ionization ultraviolet (UV) background fluctuations. A similar conclusion arises from the redshift distribution of long dark gap ($L\ge 30 ~h^{-1}\mathrm{cMpc}$) fraction. However, the model variants are still not able to reproduce the observed strong flux correlations at unusually large scales, which remains open for further investigations.

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.

Referee Report

2 major / 1 minor

Summary. The manuscript investigates the distribution of dark gaps (regions with transmitted flux below 0.05) in the Ly-α forest during the final stages of reionization (z∼5.0-6.0) using large-volume (∼1 Gpc) simulations. A suite of model variants is constructed by varying reionization completion redshift, IGM temperature, and ionizing-photon mean free path; synthetic sightlines are generated and their cumulative dark-gap distributions are compared to XQR-30 quasar spectra in Δz=0.2 bins. The authors find qualitative agreement for most models, with a z∼5.4 completion providing the closest match and short constant mean-free-path models disfavored at lower redshifts, offering qualitative support for reionization ending at z≤5.7, while acknowledging that the models fail to reproduce observed strong Ly-α flux correlations at scales ≥150 h^{-1} cMpc.

Significance. If the results hold, the work adds a useful probe of IGM conditions at the end of reionization via dark-gap statistics, leveraging large-volume simulations to sample rare long gaps (L≥30 h^{-1} cMpc) and providing direct comparison to independent observational data. The preference for a slightly later reionization completion aligns with other emerging evidence, and the large simulation volume is a clear strength for statistical robustness. The explicit acknowledgment of the mismatch with large-scale flux correlations is transparent and highlights an important open question.

major comments (2)
  1. [Abstract] Abstract: the central claim that the z∼5.4 reionization completion provides the closest match to observed dark-gap cumulative distributions rests on the assumption that varying only reionization redshift, IGM temperature, and mean free path captures the dominant physics. However, the abstract itself states that the models cannot reproduce the reported strong Ly-α flux correlations at scales ≥150 h^{-1} cMpc; this mismatch indicates that unmodeled effects (e.g., inhomogeneous reionization topology or post-reionization UV background fluctuations) could alter the predicted dark-gap statistics and shift the relative model rankings.
  2. [Results] Results (comparison to XQR-30 data): the disfavoring of short constant mean-free-path models at lower redshifts and the support for reionization extending to z≤5.7 are presented as qualitative; given the failure to match large-scale correlations, the manuscript should quantify how sensitive the dark-gap distributions are to the inclusion of additional UV fluctuation sources or inhomogeneous reionization to confirm that the three-parameter suite is sufficient for the claimed conclusions.
minor comments (1)
  1. Figure captions should explicitly list the exact parameter values (reionization redshift, temperature, mean free path) for each model variant shown in the cumulative distribution plots to improve clarity.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive comments. We have revised the manuscript to address the concerns regarding the robustness of our conclusions given the mismatch with large-scale flux correlations, while maintaining that our three-parameter exploration provides useful qualitative insights.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim that the z∼5.4 reionization completion provides the closest match to observed dark-gap cumulative distributions rests on the assumption that varying only reionization redshift, IGM temperature, and mean free path captures the dominant physics. However, the abstract itself states that the models cannot reproduce the reported strong Ly-α flux correlations at scales ≥150 h^{-1} cMpc; this mismatch indicates that unmodeled effects (e.g., inhomogeneous reionization topology or post-reionization UV background fluctuations) could alter the predicted dark-gap statistics and shift the relative model rankings.

    Authors: We agree that the mismatch with observed large-scale flux correlations indicates that our model suite does not capture all relevant physics. Our parameter variations (reionization completion redshift, IGM temperature, and mean free path) target the dominant factors expected to shape dark-gap statistics in the final stages of reionization, as motivated by prior literature. We explicitly note in the manuscript that unmodeled effects such as enhanced post-reionization UV fluctuations could allow for earlier reionization. In the revised version, we have strengthened the abstract and discussion sections to emphasize that our model rankings and support for reionization ending at z ≤ 5.7 are qualitative within the explored parameter space, and we have added a clearer statement that additional physics could alter quantitative conclusions. revision: partial

  2. Referee: [Results] Results (comparison to XQR-30 data): the disfavoring of short constant mean-free-path models at lower redshifts and the support for reionization extending to z≤5.7 are presented as qualitative; given the failure to match large-scale correlations, the manuscript should quantify how sensitive the dark-gap distributions are to the inclusion of additional UV fluctuation sources or inhomogeneous reionization to confirm that the three-parameter suite is sufficient for the claimed conclusions.

    Authors: We concur that a quantitative assessment of sensitivity to additional UV fluctuations or full inhomogeneous reionization topology would strengthen the work. However, such an analysis requires new, computationally expensive simulations that incorporate these effects self-consistently, which lies beyond the scope of the present study. Our current suite isolates the impact of the three key parameters on dark gaps using large-volume runs that enable sampling of rare long gaps. In the revised results section, we have expanded the discussion to stress the qualitative nature of the conclusions, reiterated that the models cannot rule out early reionization with enhanced UV fluctuations, and highlighted the need for future work with more comprehensive modeling to test robustness. revision: partial

Circularity Check

0 steps flagged

Direct simulation-to-observation comparison with no fitted predictions or self-definitional steps

full rationale

The paper constructs a suite of large-volume reionization simulations by varying three physical parameters (reionization redshift, IGM temperature, ionizing-photon mean free path) to generate synthetic Ly-α sightlines. It then computes the cumulative distribution of dark gaps (regions with transmitted flux below 0.05) at Δz=0.2 intervals and compares these directly to independent XQR-30 quasar spectra. No equation defines a model quantity in terms of the observed dark-gap statistics, no parameter is fitted to the target data, and no self-citation is invoked as a load-bearing uniqueness theorem. The analysis explicitly notes that the models fail to reproduce observed large-scale flux correlations, confirming the comparison remains external and non-circular.

Axiom & Free-Parameter Ledger

3 free parameters · 2 axioms · 0 invented entities

The central claim rests on the assumption that the chosen simulation framework and three-parameter variations adequately represent the physical conditions affecting dark gaps during reionization end stages.

free parameters (3)
  • reionization completion redshift
    Varied across model suite to identify best match to observed dark-gap distributions
  • IGM temperature
    Varied as one of the three main model parameters
  • ionizing-photon mean free path
    Varied; short constant value disfavored at lower redshifts
axioms (2)
  • standard math Standard flat Lambda-CDM cosmology governs the large-scale structure evolution
    Underlying assumption for the simulation volume and initial conditions
  • domain assumption Reionization is driven by UV photons from galaxies and quasars with a spatially varying background
    Core physical model for generating the ionization field in the simulations

pith-pipeline@v0.9.0 · 5650 in / 1620 out tokens · 68858 ms · 2026-05-16T08:52:01.182177+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

49 extracted references · 49 canonical work pages

  1. [1]

    S., Bell, J

    Almgren, A. S., Bell, J. B., Lijewski, M. J., Luki ´c, Z., & Van Andel, E. 2013, ApJ, 765, 39 Bañados, E., Venemans, B. P., Mazzucchelli, C., et al. 2018, Nature, 553, 473

    work page 2013

  2. [2]

    & Loeb, A

    Barkana, R. & Loeb, A. 2001, Phys. Rep., 349, 125

    work page 2001

  3. [3]

    D., Bolton, J

    Becker, G. D., Bolton, J. S., Madau, P., et al. 2015, MNRAS, 447, 3402

    work page 2015

  4. [4]

    D., D’Aloisio, A., Christenson, H

    Becker, G. D., D’Aloisio, A., Christenson, H. M., et al. 2021, MNRAS, 508, 1853

    work page 2021

  5. [5]

    Bosman, S. E. I., Davies, F. B., Becker, G. D., et al. 2022, MNRAS, 514, 55

    work page 2022

  6. [6]

    Bosman, S. E. I., Davies, F. B., Becker, G. D., et al. 2021, arXiv e-prints, arXiv:2108.03699

    work page arXiv 2021

  7. [7]

    Choudhury, T. R. 2009, Current Science, 97, 841

    work page 2009

  8. [8]

    Choudhury, T. R. 2022, General Relativity and Gravitation, 54, 102

    work page 2022

  9. [9]

    R., Paranjape, A., & Bosman, S

    Choudhury, T. R., Paranjape, A., & Bosman, S. E. I. 2021, MNRAS, 501, 5782

    work page 2021

  10. [10]

    M., Becker, G

    Christenson, H. M., Becker, G. D., D’Aloisio, A., et al. 2023, ApJ, 955, 138 D’Aloisio, A., McQuinn, M., & Trac, H. 2015, ApJ, 813, L38

    work page 2023

  11. [11]

    B., Bosman, S

    Davies, F. B., Bosman, S. E. I., D’Odorico, V ., et al. 2025, MN- RAS[arXiv:2510.25829]

    work page arXiv 2025

  12. [12]

    B., Bosman, S

    Davies, F. B., Bosman, S. E. I., Gaikwad, P., et al. 2023, arXiv e-prints, arXiv:2312.08464

    work page arXiv 2023

  13. [13]

    B., Bosman, S

    Davies, F. B., Bosman, S. E. I., Gaikwad, P., et al. 2024, ApJ, 965, 134

    work page 2024

  14. [14]

    Davies, F. B. & Furlanetto, S. R. 2016, MNRAS, 460, 1328

    work page 2016

  15. [15]

    B., Hennawi, J

    Davies, F. B., Hennawi, J. F., Bañados, E., et al. 2018, ApJ, 864, 142

    work page 2018

  16. [16]

    & Ferrara, A

    Dayal, P. & Ferrara, A. 2018, Phys. Rep., 780, 1 D’Odorico, V ., Bañados, E., Becker, G. D., et al. 2023, MNRAS, 523, 1399

    work page 2018

  17. [17]

    A., Becker, R

    Fan, X., Strauss, M. A., Becker, R. H., et al. 2006, AJ, 132, 117

    work page 2006

  18. [18]

    G., Davies, F

    Gaikwad, P., Haehnelt, M. G., Davies, F. B., et al. 2023, MNRAS, 525, 4093

    work page 2023

  19. [19]

    G., et al

    Gaikwad, P., Rauch, M., Haehnelt, M. G., et al. 2020, MNRAS, 494, 5091

    work page 2020

  20. [20]

    Gallerani, S., Ferrara, A., Fan, X., & Choudhury, T. R. 2008, MNRAS, 386, 359

    work page 2008

  21. [21]

    Y ., Becker, G

    Gnedin, N. Y ., Becker, G. D., & Fan, X. 2017, ApJ, 841, 26

    work page 2017

  22. [22]

    Gnedin, N. Y . & Madau, P. 2022, arXiv e-prints, arXiv:2208.02260

    work page arXiv 2022

  23. [23]

    2019, MNRAS, 484, 5094

    Greig, B., Mesinger, A., & Bañados, E. 2019, MNRAS, 484, 5094

    work page 2019

  24. [24]

    2024, MNRAS, 530, 3208

    Greig, B., Mesinger, A., Bañados, E., et al. 2024, MNRAS, 530, 3208

    work page 2024

  25. [25]

    B., et al

    Greig, B., Mesinger, A., Davies, F. B., et al. 2022, MNRAS, 512, 5390

    work page 2022

  26. [26]

    2019, ApJ, 878, 12

    Hoag, A., Bradaˇc, M., Huang, K., et al. 2019, ApJ, 878, 12

    work page 2019

  27. [27]

    2023, ApJ, 942, 59

    Jin, X., Yang, J., Fan, X., et al. 2023, ApJ, 942, 59

    work page 2023

  28. [28]

    C., Bunker, A

    Jones, G. C., Bunker, A. J., Saxena, A., et al. 2025, MNRAS, 536, 2355

    work page 2025

  29. [29]

    2025, ApJS, 278, 33

    Kageura, Y ., Ouchi, M., Nakane, M., et al. 2025, ApJS, 278, 33

    work page 2025

  30. [30]

    J., Shibuya, T., Ouchi, M., & Kashikawa, N

    Kashino, D., Lilly, S. J., Shibuya, T., Ouchi, M., & Kashikawa, N. 2020, ApJ, 888, 6

    work page 2020

  31. [31]

    C., Kulkarni, G., Haehnelt, M

    Keating, L. C., Kulkarni, G., Haehnelt, M. G., Chardin, J., & Aubert, D. 2020, MNRAS, 497, 906

    work page 2020

  32. [32]

    C., Haehnelt, M

    Kulkarni, G., Keating, L. C., Haehnelt, M. G., et al. 2019, MNRAS, 485, L24

    work page 2019

  33. [33]

    2025, arXiv e-prints, arXiv:2505.09369

    Maity, B., Davies, F., & Gaikwad, P. 2025, arXiv e-prints, arXiv:2505.09369

    work page arXiv 2025

  34. [34]

    A., Treu, T., Dijkstra, M., et al

    Mason, C. A., Treu, T., Dijkstra, M., et al. 2018, ApJ, 856, 2

    work page 2018

  35. [35]

    2016, ARA&A, 54, 313

    McQuinn, M. 2016, ARA&A, 54, 313

    work page 2016

  36. [36]

    M., Mason, C

    Morales, A. M., Mason, C. A., Bruton, S., et al. 2021, ApJ, 919, 120

    work page 2021

  37. [37]

    & D’Aloisio, A

    Nasir, F. & D’Aloisio, A. 2020, MNRAS, 494, 3080 Planck Collaboration, Aghanim, N., Akrami, Y ., et al. 2020, A&A, 641, A6

    work page 2020

  38. [38]

    2025, PASA, 42, e049

    Qin, Y ., Mesinger, A., Prelogovi´c, D., et al. 2025, PASA, 42, e049

    work page 2025

  39. [39]

    & Mesinger, A

    Sobacchi, E. & Mesinger, A. 2015, MNRAS, 453, 1843

    work page 2015

  40. [40]

    Spina, B., Bosman, S. E. I., Davies, F. B., Gaikwad, P., & Zhu, Y . 2024, A&A, 688, L26 Article number, page 8 of 10 Maity et al.: Studying dark gaps in Ly-αforest transmission with large reionization simulation

    work page 2024

  41. [41]

    Spina, B., Bosman, S. E. I., Maity, B., & Davies, F. B. 2025, arXiv e-prints, arXiv:2507.12170

    work page arXiv 2025

  42. [42]

    A., & Cen, R

    Trac, H., Chen, N., Holst, I., Alvarez, M. A., & Cen, R. 2022, ApJ, 927, 186

    work page 2022

  43. [43]

    Umeda, M

    Umeda, H., Ouchi, M., Kageura, Y ., et al. 2025a, arXiv e-prints, arXiv:2504.04683

    work page arXiv

  44. [44]

    2024, ApJ, 971, 124

    Umeda, H., Ouchi, M., Nakajima, K., et al. 2024, ApJ, 971, 124

    work page 2024

  45. [45]

    2011, A&A, 536, A105

    Vernet, J., Dekker, H., D’Odorico, S., et al. 2011, A&A, 536, A105

    work page 2011

  46. [46]

    B., Yang, J., et al

    Wang, F., Davies, F. B., Yang, J., et al. 2020, ApJ, 896, 23 Zel’dovich, Y . B. 1970, A&A, 5, 84

    work page 2020

  47. [47]

    D., Bosman, S

    Zhu, Y ., Becker, G. D., Bosman, S. E. I., et al. 2024, MNRAS, 533, L49

    work page 2024

  48. [48]

    D., Bosman, S

    Zhu, Y ., Becker, G. D., Bosman, S. E. I., et al. 2021, ApJ, 923, 223

    work page 2021

  49. [49]

    late", "early

    Zhu, Y ., Becker, G. D., Christenson, H. M., et al. 2023, ApJ, 955, 115 Article number, page 9 of 10 A&A proofs:manuscript no. aanda 5 5.2 5.4 5.6 5.8 6 Redshift (z) late 5 5.2 5.4 5.6 5.8 6 Redshift (z) early 5 5.2 5.4 5.6 5.8 6 Redshift (z) small mfp 5 5.2 5.4 5.6 5.8 6 Redshift (z) large mfp 500 1000 h 1cMpc 5 5.2 5.4 5.6 5.8 6 Redshift (z) 500 1000 h ...

    work page 2023