[Paper Review] Dark energy survey year 3 results: Cosmology with peaks using an emulator approach

D. Zürcher, Janis Fluri+120 more|arXiv (Cornell University)|Jan 1, 2022

Cosmology and Gravitation TheoriesPhysics and Astronomy131 references70 citations

TL;DR

This paper presents cosmological constraints from Dark Energy Survey Year 3 weak lensing data using tomographic shear peak counts and angular convergence power spectra, employing an N-body simulation-based Gaussian process emulator to model non-Gaussian statistics. It achieves a 1.8% precision on 𝑆8 = 𝜎8√(Ωm/0.3) = 0.797+0.015−0.013, representing a 38% improvement over power spectra alone, with robustness confirmed through systematic tests and scale cuts to mitigate baryonic physics biases.

ABSTRACT

We constrain the matter density ωm and the amplitude of density fluctuations σ8 within the ΛCDM cosmological model with shear peak statistics and angular convergence power spectra using mass maps constructed from the first three years of data of the Dark Energy Survey (DES Y3). We use tomographic shear peak statistics, including cross-peaks: peak counts calculated on maps created by taking a harmonic space product of the convergence of two tomographic redshift bins. Our analysis follows a forward-modelling scheme to create a likelihood of these statistics using N-body simulations, using a Gaussian process emulator. We take into account the uncertainty from the remaining, largely unconstrained ΛCDM parameters (ωb, ns, and h). We include the following lensing systematics: multiplicative shear bias, photometric redshift uncertainty, and galaxy intrinsic alignment. Stringent scale cuts are applied to avoid biases from unmodelled baryonic physics. We find that the additional non-Gaussian information leads to a tightening of the constraints on the structure growth parameter yielding S8 σ8√Ωm/0.3=0.797-0.013+0.015 (68 per cent confidence limits), with a precision of 1.8 per cent, an improvement of 38 per cent compared to the angular power spectra only case. The results obtained with the angular power spectra and peak counts are found to be in agreement with each other and no significant difference in S8 is recorded. We find a mild tension of 1.5 σ between our study and the results from Planck 2018, with our analysis yielding a lower S8. Furthermore, we observe that the combination of angular power spectra and tomographic peak counts breaks the degeneracy between galaxy intrinsic alignment AIA and S8, improving cosmological constraints. We run a suite of tests concluding that our results are robust and consistent with the results from other studies using DES Y3 data.

Motivation & Objective

• To improve cosmological constraints on Ωm and 𝑆8 by extracting non-Gaussian information from shear peak counts in addition to angular power spectra.
• To develop a forward-modelling likelihood using a Gaussian process emulator trained on N-body simulations to efficiently predict peak counts across cosmological parameter space.
• To account for uncertainties in ΛCDM parameters (Ωb, 𝑛s, ℎ) and major lensing systematics: multiplicative shear bias, photometric redshift errors, and intrinsic alignments.
• To mitigate biases from unmodelled baryonic physics by applying scale cuts derived from baryon-inclusive simulations, ensuring results remain within 0.3𝜎 of the true cosmology.
• To test the robustness of constraints through null tests (B-mode), systematic checks, and consistency with other DES Y3 analyses.

Proposed method

• Uses a forward-modelling approach: simulate weak lensing mass maps from N-body simulations at various cosmologies to predict peak counts and angular power spectra.
• Employs a Gaussian process emulator trained on a grid of cosmological simulations to interpolate the statistics across the full parameter space (Ωm, 𝜎8, Ωb, 𝑛s, ℎ) efficiently.
• Combines tomographic peak counts (including cross-peaks between redshift bins) with angular power spectra to enhance cosmological sensitivity.
• Applies stringent scale cuts (ℓ∈[30, 578], FWHM∈[7.9, 31.6] arcmin) derived from baryon-contaminated simulations to limit bias from missing baryonic feedback effects.
• Uses synthetic data vectors from baryon-inclusive simulations (via the Schneider et al. 2019 model) to estimate and correct for baryonic systematics, ensuring shifts in constraints remain ≤0.3𝜎.
• Performs null tests on B-mode power spectra and checks for cosmology-dependent biases in the marginalisation scheme, confirming robustness.

Experimental results

Research questions

• RQ1 Can tomographic shear peak counts combined with angular power spectra improve cosmological constraints on 𝑆8 beyond what is achievable with power spectra alone?
• RQ2 How do baryonic physics effects, particularly feedback-induced halo profile changes, bias peak count predictions, and what scale cuts are needed to control this?
• RQ3 To what extent do systematic effects—such as multiplicative shear bias, photometric redshift uncertainty, and intrinsic alignments—affect the final cosmological constraints?
• RQ4 Does the combination of peak counts and power spectra break degeneracies between 𝑆8 and the intrinsic alignment amplitude 𝐴IA, improving overall constraints?
• RQ5 Are the results robust under null tests (e.g., B-mode power) and systematic checks, including cosmology-dependent marginalisation and boost factor corrections?

Key findings

• The combination of angular power spectra and tomographic peak counts yields a 38% tighter constraint on 𝑆8 compared to power spectra alone, achieving a precision of 1.8%: 𝑆8 = 0.797+0.015−0.013 (68% CL).
• The results from peak counts and power spectra are in excellent agreement, with no significant discrepancy in the inferred 𝑆8 value.
• A mild 1.5𝜎 tension is observed with Planck 2018 results, with this analysis yielding a lower 𝑆8 value.
• The inclusion of peak counts breaks the degeneracy between 𝑆8 and the intrinsic alignment amplitude 𝐴IA, improving the overall cosmological sensitivity.
• Scale cuts of ℓ∈[30, 578] and FWHM∈[7.9, 31.6] arcmin limit the bias from missing baryonic physics to ≤0.3𝜎, satisfying the unblinding criteria.
• Null tests on B-mode power spectra pass with p > 1% (p = 31% for peaks, p = 23% for power spectra), confirming no significant residual systematics.