[Paper Review] DES Y3 cosmic shear down to small scales: constraints on cosmology and baryons

Giovanni Aricò, Raúl E. Angulo+4 more|arXiv (Cornell University)|Mar 9, 2023

Galaxies: Formation, Evolution, Phenomena6 citations

TL;DR

This study presents the first DES Y3 cosmic shear analysis to include all angular scales down to small scales by leveraging advanced modeling of baryonic physics via BACCOemu, an emulator combining N-body simulations and neural networks. It tightens constraints on cosmological parameters, measuring $S_8 = 0.799^{+0.023}_{-0.015}$ and $\log M_c = 14.38^{+0.60}_{-0.56}$, with results consistent with Planck at 0.9σ but showing a 1.4σ shift relative to the official DES Y3 shear analysis due to improved modeling of baryons, non-linearities, and intrinsic alignments.

ABSTRACT

We present the first analysis of cosmic shear measured in DES Y3 that employs the entire range of angular scales in the data. To achieve this, we build upon recent advances in the theoretical modelling of weak lensing provided by a combination of $N$-body simulations, physical models of baryonic processes, and neural networks. Specifically, we use BACCOemu to model the linear and nonlinear matter power spectrum including baryonic physics, allowing us to robustly exploit scales smaller than those used by the DES Collaboration. We show that the additional data produce cosmological parameters that are tighter but consistent with those obtained from larger scales, while also constraining the distribution of baryons. In particular, we measure the mass scale at which haloes have lost half of their gas, $\log\,M_{ m c}=14.38^{+0.60}_{-0.56}\log(h^{-1}{ m M_{ \odot}})$, and a parameter that quantifies the weighted amplitudes of the present-day matter inhomogeneities, $S_8=0.799^{+0.023}_{-0.015}$. Our constraint on $S_8$ is statistically compatible with that inferred from the Planck satellite's data at the $0.9σ$ level. We find instead a $1.4σ$ shift in comparison to that from the official DES Y3 cosmic shear, because of different choices in the modelling of intrinsic alignment, non-linearities, baryons, and lensing shear ratios. We conclude that small scales in cosmic shear data contain valuable astrophysical and cosmological information and thus should be included in standard analyses.

Motivation & Objective

• To reanalyze DES Y3 cosmic shear data using the full range of angular scales, including small scales previously excluded due to modeling challenges.
• To investigate whether baryonic physics—particularly feedback-driven gas redistribution—can alleviate the $S_8$ tension between CMB and weak lensing measurements.
• To develop and apply a robust, physically motivated model of the matter power spectrum that includes non-linear and baryonic effects using $N$-body simulations and neural network emulators.
• To constrain the halo gas mass scale $M_c$ where haloes lose half their gas, providing a direct probe of baryonic feedback efficiency.
• To assess the impact of improved modeling of intrinsic alignments, shear ratios, and non-linearities on cosmological parameter constraints.

Proposed method

• Utilized BACCOemu, a neural network-based emulator trained on $N$-body simulations, to model the linear and nonlinear matter power spectrum including baryonic physics across cosmological parameters.
• Extended the cosmic shear analysis to small scales (high $\ell$) by incorporating physical models of baryonic feedback, such as supernova and AGN feedback, which suppress power on small scales.
• Performed a 3x2pt analysis combining cosmic shear, galaxy-galaxy lensing, and galaxy clustering, using non-linear bias emulators from BACCOemu to model galaxy clustering and lensing signals.
• Incorporated systematic effects such as intrinsic alignments and shear ratio corrections using physically motivated parametric models.
• Constrained cosmological parameters via Bayesian inference using the CosmoSIS framework, with likelihoods informed by the BACCOemu predictions.
• Validated results against Planck CMB data and the official DES Y3 cosmic shear results to assess consistency and shifts due to modeling choices.

Experimental results

Research questions

• RQ1 To what extent can baryonic feedback processes suppress small-scale power in the cosmic shear signal, and can this explain part of the $S_8$ tension?
• RQ2 How do improved models of non-linear structure growth and intrinsic alignments affect the inferred value of $S_8$ in DES Y3 cosmic shear data?
• RQ3 What constraints can be placed on the characteristic halo mass scale $M_c$ where gas loss reaches 50%, as a probe of baryonic feedback efficiency?
• RQ4 How do the results from including small-scale cosmic shear data compare to the official DES Y3 cosmic shear analysis, and what drives the observed shifts?
• RQ5 Can a unified modeling framework that includes baryons, non-linearities, and intrinsic alignments reduce systematic uncertainties in weak lensing cosmology?

Key findings

• The inclusion of small-scale cosmic shear data leads to tighter cosmological constraints, with $S_8 = 0.799^{+0.023}_{-0.015}$, consistent with Planck's $S_8$ at the 0.9σ level.
• The analysis constrains the halo gas mass scale at which haloes lose half their gas to $\log M_c = 14.38^{+0.60}_{-0.56} \, \log(h^{-1} M_\odot)$, providing a direct probe of baryonic feedback.
• The $S_8$ constraint shows a 1.4σ shift compared to the official DES Y3 cosmic shear analysis, primarily due to differences in modeling intrinsic alignments, non-linearities, baryons, and shear ratios.
• The results demonstrate that small-scale cosmic shear contains valuable cosmological and astrophysical information, and should be included in standard analyses to reduce systematic biases.
• The BACCOemu-based modeling framework successfully captures the non-linear and baryonic effects on the matter power spectrum, enabling robust small-scale inference.
• The study confirms that baryonic physics can suppress small-scale power, but this suppression alone is insufficient to resolve the $S_8$ tension, suggesting other systematics or new physics may be involved.