Cosmological constraints from galaxy clustering and galaxy-galaxy lensing with extended SubHalo Abundance Matching

Abstract

We present the first cosmological constraints from a joint analysis of galaxy clustering and galaxy-galaxy lensing using extended SubHalo Abundance Matching (SHAMe). We analyse stellar mass-selected Galaxy And Mass Assembly (GAMA) galaxy clustering and Kilo-Degree Survey (KiDS-1000) galaxy-galaxy lensing and find constraints on $S_8\equivσ_8\sqrt{Ω_{\rm m}/0.3}=0.793^{+0.025}_{-0.024}$, in agreement with Planck at 1.7$σ$, with $σ_8$ the mass density fluctuation amplitude in 8 $h^{-1}{\rm Mpc}$ sphere at present and $Ω_{\rm m}$ the density parameter in total matter. These results are in agreement with the Cosmic Microwave Background results from Planck. We are able to constrain all 5 SHAMe parameters, which describe the galaxy-subhalo connection. We validate our methodology by first applying it to simulated catalogues, generated from the TNG300 simulation, which mimic the stellar mass selection of our real data. We show that we are able to recover the input cosmology for both our fiducial and all-scale analyses. Our all-scale analysis extends to scales of galaxy-galaxy lensing below $r_\mathrm{p}<1.4\,\mathrm{Mpc}/h$, which we exclude in our fiducial analysis to avoid baryonic effects. When including all scales, we find a value of $S_8$, which is 1.26$σ$ higher than our fiducial result (against naive expectations where baryonic feedback should lead to small-scale power suppression), and in agreement with Planck at 0.9$σ$. We also find a 21% tighter constraint on $S_8$ and a 29% tighter constraint on $Ω_\mathrm{m}$ compared to our fiducial analysis. This work shows the power and potential of joint small-scale galaxy clustering and galaxy-galaxy lensing analyses using SHAMe.

Figures (10)

• Figure 1: Differential stellar mass function from TNG300 compared to GAMA Driver:2022vyh and KiDS-1000 Dvornik:2022xap. The dashed lines indicate the stellar mass bin boundaries, which are required as inputs to the SHAMe emulator. We do not include the bins in the grey-shaded region, as they fall outside the range of the emulator.
• Figure 2: TNG300 galaxy clustering (top) and TNG300 galaxy-galaxy lensing (bottom) for each of the four stellar mass bins used in this work. The solid blue line shows the best fit when analysing the joint data vector with SHAMe, where the galaxy-galaxy lensing data in the grey shaded region is not included, and the green dashed line shows the best fit when all scales of the galaxy-galaxy lensing are included in the data vector.
• Figure 3: Marginalised posterior constraints on $S_8$ and $\Omega_\mathrm{m}$ for TNG300 galaxy clustering and galaxy-galaxy lensing using SHAMe (blue) and the same analysis without scale cuts in the galaxy-galaxy lensing (green). We recover the TNG300 input cosmology (dashed lines) within $1\sigma$, where the covariance matches the covariance used in the real data analysis.
• Figure 4: Marginalised posterior constraints on $S_8$ and $\Omega_\mathrm{m}$ for GAMA galaxy clustering and KiDS-1000 galaxy-galaxy lensing using SHAMe (blue) and the same analysis without scale cuts in the galaxy-galaxy lensing (green). For comparison, we show CMB results from Planck (Planck, orange dashed) and results for KiDS-1000 galaxy clustering and galaxy-galaxy lensing using an analytic conditional stellar mass function to describe the galaxy-halo connection (Dvornik:2022xap, red dashed). The results are consistent, with SHAMe (this work) giving a slightly higher value of $S_8$, particularly when all scales are included in the galaxy-galaxy lensing.
• Figure 5: GAMA galaxy clustering (top) and KiDS-1000 galaxy-galaxy lensing (bottom) for each of the four stellar mass bins used in this work. The solid blue line shows the best fit when analysing the joint data vector with SHAMe, where the galaxy-galaxy lensing data in the grey shaded region is not included in order to avoid the impact of baryonic effects (Figure \ref{['fig:S8_omegam_all_scales_kids']}, blue contour). The green dashed line shows the best fit when all scales of the galaxy-galaxy lensing are included in the data vector (Figure \ref{['fig:S8_omegam_all_scales_kids']}, green contour).