Consistency of effective field theory analyses of the BOSS power spectrum Article Swipe
YOU?
·
· 2023
· Open Access
·
· DOI: https://doi.org/10.1103/physrevd.107.123530
· OA: W4293583710
We assess the robustness of $\\Lambda$CDM results from the full-shape analysis\nof BOSS power spectrum using the one-loop prediction of the Effective Field\nTheory of Large-Scale Structure (EFTofLSS). The public likelihoods PyBird and\nCLASS-PT lead to results in agreement only at the $1\\sigma$ level, despite the\nfact that they are derived from the same BOSS dataset and theory model. We\nperform a thorough comparison of the various analyses choices made between the\ntwo pipelines, and identify that the differences come from the choice of prior\non the EFT parameters, dubbed "West-coast" (WC) and "East-coast" (EC) prior,\nrespectively associated to PyBird and CLASS-PT. In particular, because\nposteriors are non-Gaussian, projection effects from the marginalization over\nthe EFT parameters shift the posterior mean of the cosmological parameters with\nrespect to the best-fit up to $1\\sigma$ in the WC prior and up to $2\\sigma$ in\nthe EC prior. We quantify that best-fit cosmological parameters extracted from\nBOSS given the two prior choices are consistent at $\\sim 1\\sigma$. The\nconsistency improves to $\\sim 0.5\\sigma$ when doubling the prior widths. While\nthis reveals that current EFT analyses are subject to prior effects, we show\nthat cosmological results obtained in combination with CMB, or from forthcoming\nlarge-volume data, are less sensitive to those effects. In addition, we\ninvestigate differences between BOSS measurements. We find broad agreements\nacross all pre-reconstructed measurements considered ($<0.6\\sigma$), but the\ntwo available BOSS post-reconstructed measurements in Fourier space, once\ncombined with the EFT full-shape analysis, lead to discrepant Hubble parameter\n$H_0$ at $\\sim 0.9\\sigma$. Given the various effects we discuss, we argue that\nthe clustering amplitude $\\sigma_8$ measured with BOSS is not in statistical\ntension with that inferred from Planck under $\\Lambda$CDM.\n
