CosmoLike - Cosmological Likelihood Analyses for Photometric Galaxy Surveys

TL;DR

This work develops and applies CosmoLike to forecast cosmological constraints from a joint analysis of multiple photometric probes (cosmic shear, galaxy–galaxy lensing, galaxy clustering, photometric BAOs, cluster counts, and cluster lensing) for an LSST-like survey. It incorporates cross-probe correlations and non-Gaussian covariances, modeling a wide set of systematics and up to 54 nuisance parameters, to quantify information content as a function of scale and modeling assumptions. The key findings show that while increasing source density yields diminishing returns under current systematics, including small-scale clustering via HOD modeling significantly boosts constraints, and that the full multi-probe combination can dramatically tighten cosmological parameters compared with single-probe analyses. The results inform survey design and motivate robust, cross-probe modeling of systematics, with CosmoLike made publicly available for broader use and extension.

Abstract

We explore strategies to extract cosmological constraints from a joint analysis of cosmic shear, galaxy-galaxy lensing, galaxy clustering, cluster number counts and cluster weak lensing. We utilize the CosmoLike software to simulate results from an LSST like data set, specifically, we 1) compare individual and joint analyses of the different probes, 2) vary the selection criteria for lens and source galaxies, 3) investigate the impact of blending, 4) investigate the impact of the assumed cosmological model in multi-probe covariances, 6) quantify information content as a function of scales, and 7) explore the impact of intrinsic galaxy alignment in a multi-probe context. Our analyses account for all cross correlations within and across probes and include the higher-order (non-Gaussian) terms in the multi-probe covariance matrix. We simultaneously model cosmological parameters and a variety of systematics, e.g. uncertainties arising from shear and photo-z calibration, cluster mass-observable relation, galaxy intrinsic alignment, and galaxy bias (up to 54 parameters altogether). We highlight two results: First, increasing the number density of source galaxies by ~30%, which corresponds to solving blending for LSST, only gains little information. Second, including small scales in clustering and galaxy-galaxy lensing, by utilizing HODs, can substantially boost cosmological constraining power. The CosmoLike modules used to compute the results in this paper will be made publicly available at https://github.com/elikrause/CosmoLike_Forecasts.

Figures (6)

• Figure 1: Non-Gaussian, multi-probe correlation matrix for a joint data vector of cosmic shear, galaxy-galaxy lensing, galaxy clustering, cluster number counts, and cluster weak lensing. Details on the calculation of this $\sim$ 5.8 million entry matrix (7.4 million when going to the smallest scales considered in this paper) can be found in Appendix \ref{['sec:app1']}. We recommend a zoom factor of $\sim$10 to gain more insight into the matrix structure, e.g. to actually identify individual elements.
• Figure 2: Individual vs. multi-probe cosmological constraints. We show projected cosmological constraints for clustering (orange/dot-long dashed), cosmic shear (red/dashed), cluster number counts (blue/dot-dashed) individually. The 3x2pt multi-probe contours (green/long-dashed) include information from clustering, cosmic shear, and galaxy-galaxy lensing; the black/solid contours add information from cluster number counts and cluster weak lensing to the 3x2pt data vector, altogether 2413 data points.
• Figure 3: Impact of galaxy samples and associate systematics on cosmological information. We show the systematics free 3x2pt function case (black, solid) in comparison to our baseline model (red/dashed). The (blue, dot-dashed) contours show the information gain when including all blended objects in the analysis, i.e. increasing $\bar{n}_\mathrm{source}$ from 26 to 37 galaxies/arcmin$^2$; green/long-dashed constraints are obtained when including a lens galaxy sample that is by a factor of 20 larger than our baseline (red sequence) sample, but has worse photo-z accuracy.
• Figure 4: Left: Varying the minimum scale included in galaxy clustering and galaxy galaxy lensing measurements. We show the baseline 3x2pt functions, which assumes $R_\mathrm{min}=10\mathrm{ Mpc/h}$(black/solid), and corresponding constraints when using $R_\mathrm{min}=20 \mathrm{Mpc/h}$(red/dashed), $R_\mathrm{min}=50 \mathrm{Mpc/h}$(blue/dot-dashed), $R_\mathrm{min}=0.1 \mathrm{Mpc/h}$(green/long-dashed) instead. For the latter we switch from linear galaxy bias modeling to our HOD implementation. Right: Information gain when using HOD instead of linear galaxy bias for 3x2pt (black solid vs dashed contours) in comparison to corresponding information gain when including cluster number counts and cluster weak lensing in the data vector (violett/dot-dashed vs long-dashed).
• Figure 5: Change in cosmological constraints when varying the underlying cosmological model in the covariance matrix. We show three scenarios: 1) the fiducial cosmology (black/solid), 2) fiducial cosmology but a 10% lower value in $\sigma_8$ and $\Omega_\mathrm m$(red/dashed), and 3) fiducial cosmology but changes in the dark energy parameters, i.e. $w_0=-1.3$ and $w_a=-0.5$(blue/dot-dashed).