How much information can be extracted from galaxy clustering at the field level?

Abstract

We present optimal Bayesian field-level cosmological constraints from nonlinear tracers of the large-scale structure, specifically the amplitude $\ddot{I}{ƒ}_8$ of linear matter fluctuations inferred from rest-frame simulated dark matter halos in a comoving volume of $8(h^{-1}\mathrm{Gpc}{)}^3$. Our constraint on $\ddot{I}{ƒ}_8$ is entirely due to nonlinear information, and obtained by explicitly sampling the initial conditions along with bias and noise parameters via a Lagrangian EFT-based forward model, LEFTfield. The comparison with a simulation-based inference analysis employing the power spectrum and bispectrum -- likewise using the LEFTfield forward model -- shows that, when including precisely the same modes of the same data up to $k_{\max }=0.10h{\mathrm{Mpc}}^{-1}$ ($0.12h{\mathrm{Mpc}}^{-1}$), the field-level approach yields a factor of 3.5 (5.2) improvement on the $\ddot{I}{ƒ}_8$ constraint, from 20.0% to 5.7% (17.0% to 3.3%). This study provides direct insights into cosmological information encoded in galaxy clustering beyond low-order $n$-point functions.

3750 words + 3 appendices. 17 pages (including references), 12 figures. Main results in Figs. 2 & 3. All comments welcome! v2: Less words and more results. SBI P+B results, comparisons and discussion thereof updated. SBI P+B and FBI now use exactly the same forward model, yielding completely consistent posterior means. v3: match the journal-accepted version; PRL in press