EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Tracing the cosmic origins: machine learning reconstruction of the primordial density field from EoR observations

Anchal Saxena, P Daniel Meerburg, Guochao Sun, Tzu-Ching Chang, Lluís Mas-Ribas

Monthly Notices of the Royal Astronomical Society · 05.11.2026 · 553 (2)

DOI: 10.1093/mnras/stag1972 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Monthly Notices of the Royal Astronomical Society, online erschienen am 16.10.2026. Verfasst von Anchal Saxena und 4 weiteren Autor:innen. Die Arbeit nennt 106 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: CIERA, NSF, NASA.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Reconstructing the initial conditions of the Universe from late-time tracers would unlock cosmological information buried by non-linear structure formation and astrophysics. We reconstruct the initial density field corresponding to $z=300$, with its linear amplitude extrapolated to $z=0$, from simulated 21-cm and CO(1–0) line-intensity maps at $z=8$ generated with the seminumerical code, limfast. Using a three-dimensional U-shaped convolutional neural network (U-Net), we reconstruct the initial conditions and evaluate their impact on cosmological parameter constraints. The two tracers probe complementary environments: 21-cm emission traces neutral low-density regions of the intergalactic medium, while CO traces overdense star-forming regions. We additionally train and evaluate the reconstruction on maps incorporating Square Kilometre Array Phase 1 Low (SKA1-Low)- and Carbon Monoxide Mapping Array Project-Extended Reionization Array-like instrumental response and thermal noise. In the noiseless case, combining both tracers delivers the most accurate reconstruction across ionization states, with the cross-correlation coefficient between the reconstructed and true initial density fields, $|C(k)|\gtrsim 0.90$ for $k\lesssim 0.75\, {\rm Mpc}^{-1}$. With observational effects, small-scale information is degraded, but the joint reconstruction retains substantial large-scale coherence, with $|C(k)|\gtrsim 0.70$ for $k\lesssim 0.3\, {\rm Mpc}^{-1}$. To quantify the cosmological information gain, we perform simulation-based inference using power-spectrum summaries pre- and post-reconstruction. Across an ensemble of realizations, reconstruction yields the largest improvements for $\sigma _8$ and $n_{\rm s}$, with a median aggregate marginal compression of 1.17 bits in the noiseless case, corresponding to a 56 per cent reduction in the factorized product of prior-normalized 68 per cent marginal widths. With observational effects, the corresponding median compression is 1.29 bits. These results demonstrate that joint 21-cm–CO reconstruction can recover cosmological information beyond that captured by the pre-reconstruction summaries.

Förderung

  • CIERA
  • NSF
  • NASA
  • STScI

Zitieren

APA

Saxena, A., Meerburg, P. D., Sun, G., Chang, T., & Mas-Ribas, L. (2026). Tracing the cosmic origins: machine learning reconstruction of the primordial density field from EoR observations. Monthly Notices of the Royal Astronomical Society, 553(2). https://doi.org/10.1093/mnras/stag1972

Vancouver

Saxena A, Meerburg PD, Sun G, Chang T, Mas-Ribas L. Tracing the cosmic origins: machine learning reconstruction of the primordial density field from EoR observations. Monthly Notices of the Royal Astronomical Society. 2026;553(2). doi:10.1093/mnras/stag1972

BibTeX
@article{Saxena2026,
  author = {Saxena, Anchal and Meerburg, P Daniel and Sun, Guochao and Chang, Tzu-Ching and Mas-Ribas, Lluís},
  title = {Tracing the cosmic origins: machine learning reconstruction of the primordial density field from EoR observations},
  journal = {Monthly Notices of the Royal Astronomical Society},
  year = {2026},
  volume = {553},
  number = {2},
  doi = {10.1093/mnras/stag1972}
}
Kostenlos starten →