[Paper Review] Dark Energy Survey Year 3 Results: Multi-Probe Modeling Strategy and Validation

E. Krause, Xiao Fang+122 more|arXiv (Cornell University)|May 28, 2021

Galaxies: Formation, Evolution, Phenomena45 citations

TL;DR

The paper presents the DES Year 3 modeling pipeline for a 3×2pt analysis, validates baseline choices with simulations and cross-pipeline consistency, and assesses systematics mitigation.

ABSTRACT

This paper details the modeling pipeline and validates the baseline analysis choices of the DES Year 3 joint analysis of galaxy clustering and weak lensing (a so-called "3$ imes$2pt" analysis). These analysis choices include the specific combination of cosmological probes, priors on cosmological and systematics parameters, model parameterizations for systematic effects and related approximations, and angular scales where the model assumptions are validated. We run a large number of simulated likelihood analyses using synthetic data vectors to test the robustness of our baseline analysis. We demonstrate that the DES Year 3 modeling pipeline, including the calibrated scale cuts, is sufficiently accurate relative to the constraining power of the DES Year 3 analyses. Our systematics mitigation strategy accounts for astrophysical systematics, such as galaxy bias, intrinsic alignments, source and lens magnification, baryonic effects, and source clustering, as well as for uncertainties in modeling the matter power spectrum, reduced shear, and estimator effects. We further demonstrate excellent agreement between two independently-developed modeling pipelines, and thus rule out any residual uncertainties due to the numerical implementation.

Motivation & Objective

• Define the DES-Y3 baseline modeling choices for 3×2pt analyses (cosmology, systematics, and approximations).
• Demonstrate robustness of the baseline parameterization via simulated likelihood analyses and stress tests.
• Validate mitigation of astrophysical systematics (galaxy bias, intrinsic alignments, magnification, baryonic effects) and modeling uncertainties (power spectrum, reduced shear, estimator effects).
• Show agreement between two independently developed modeling pipelines to rule out numerical implementation uncertainties.

Proposed method

• Derive angular 3×2pt statistics under a flat wCDM cosmology with leading-order lensing distortions.
• Parameterize 3D power spectra and include nonlinear structure formation and astrophysical effects (IA, bias, magnification, baryons).
• Use Limber approximation for cross-spectra with validated caveats and apply bin-averaged transforms to obtain w(θ), γt(θ), and ξ±(θ).
• Perform extensive simulated likelihood analyses with synthetic data vectors to test robustness of baseline choices and scale cuts.
• Compare outcomes from two independent modeling pipelines to verify numerical consistency.

Experimental results

Research questions

• RQ1 What baseline modeling choices are necessary for DES-Y3 3×2pt analyses to be robust against systematics?
• RQ2 Do simulated likelihood tests support the accuracy of the DES-Y3 modeling pipeline relative to its statistical power?
• RQ3 How do astrophysical and modeling systematics affect DES-Y3 cosmological inferences, and can they be mitigated effectively?
• RQ4 Are two independently developed modeling pipelines in agreement, ensuring numerical implementation does not bias results?],
• RQ5 key_findings:[