The clustering of galaxies in the SDSS-III Baryon Oscillation Spectroscopic Survey: BAO measurement from the LOS-dependent power spectrum of DR12 BOSS galaxies

TL;DR

This work delivers an anisotropic BAO analysis of the final DR12 BOSS data by fitting the monopole and μ^2-moment of the power spectrum for the LOWZ and CMASS samples, both pre- and post-reconstruction. The authors introduce a streamlined μ^2-moment fitting approach that preserves all BAO information while simplifying modelling, and they validate their method with MD-Patchy and QPM mocks to control systematics. Post-reconstruction results yield precise LOS-BAO measurements, which are combined with independent correlation-function analyses to obtain consensus constraints on H(z)r_s(z_d) and D_A(z)/r_s(z_d) at z≈0.32 and z≈0.57, with D_V(z)/r_s(z_d) also derived. The findings demonstrate minimal systematic biases and demonstrate that LOS BAO measurements from the power spectrum are consistent with, and competitive with, configuration-space analyses, enhancing constraints on cosmic expansion history.

Abstract

[abridged] We present an anisotropic analysis of the baryonic acoustic oscillation (BAO) scale in the twelfth and final data release of the Baryonic Oscillation Spectroscopic Survey (BOSS). We independently analyse the LOWZ and CMASS galaxy samples: the LOWZ sample contains contains 361 762 galaxies with an effective redshift of $z_{\rm LOWZ}=0.32$; the CMASS sample consists of 777 202 galaxies with an effective redshift of $z_{\rm CMASS}=0.57$. We extract the BAO peak position from the monopole power spectrum moment, $α_0$, and from the $μ^2$ moment, $α_2$, where $μ$ is the cosine of the angle to the line-of-sight. The $μ^2$-moment provides equivalent information to that available in the quadrupole but is simpler to analyse. After applying a reconstruction algorithm to reduce the BAO suppression by bulk motions, we measure the BAO peak position in the monopole and $μ^2$-moment, which are related to radial and angular shifts in scale. We report $H(z_{\rm LOWZ})r_s(z_d)=(11.60\pm0.60)\cdot10^3 {\rm km}s^{-1}$ and $D_A(z_{\rm LOWZ})/r_s(z_d)=6.66\pm0.16$ with a cross-correlation coefficient of $r_{HD_A}=0.41$, for the LOWZ sample; and $H(z_{\rm CMASS})r_s(z_d)=(14.56\pm0.37)\cdot10^3 {\rm km}s^{-1}$ and $D_A(z_{\rm CMASS})/r_s(z_d)=9.42\pm0.13$ with a cross-correlation coefficient of $r_{HD_A}=0.47$, for the CMASS sample. We combine these results with the measurements of the BAO peak position in the monopole and quadrupole correlation function of the same dataset \citep[][companion paper]{Cuestaetal2015} and report the consensus values: $H(z_{\rm LOWZ})r_s(z_d)=(11.63\pm0.69)\cdot10^3 {\rm km}s^{-1}$ and $D_A(z_{\rm LOWZ})/r_s(z_d)=6.67\pm0.15$ with $r_{HD_A}=0.35$ for the LOWZ sample; $H(z_{\rm CMASS})r_s(z_d)=(14.67\pm0.42)\cdot10^3 {\rm km}s^{-1}$ and $D_A(z_{\rm CMASS})/r_s(z_d)=9.47\pm0.12$ with $r_{HD_A}=0.52$ for the CMASS sample.

Figures (3)

• Figure 1: The measured LOWZ (top panel) and CMASS (bottom panel) DR12 post-recon, monopole (blue squares), quadrupole (red circles) and $\mu^2$-moment (green triangles) power spectra. For all the cases the measurements correspond to a combination of the northern and southern galaxy caps according to their effective areas as described in §\ref{['sec:estimator_ps']}. The error-bars are calculated from the dispersion of measurements using the qpm mocks. The red, blue and green lines correspond to the best-fitting model of Eq. (\ref{['Pmodel']}) with the BAO peak position as a free parameter. Within each panel we also present the power spectrum monopole and $\mu^2$-moment divided by the smooth power spectrum calculated in our fit to the data. For the monopole and $\mu^2$-moment we see how the model is able to capture the BAO features observed in the data. For clarity, the bottom sub-panels show the residuals between the measurement and the model, $\Delta P\equiv P^{\rm model} - P^{\rm data}$, divided by the $1\sigma$ error of the data. The black dashed lines marks the $1\sigma$ and $2\sigma$ deviations.
• Figure 2: Likelihood surfaces in the $\alpha_0-\alpha_2$ parameter space obtained from measuring the power spectrum DR12 data monopole and $\mu^2$-moment. The black points are the (down-sampled) output of the MCMC chain, for the CMASS and LOWZ samples, for both pre- and post-reconstructed catalogues, as labeled. In all cases the qpm-covariance matrix has been used for estimating the power spectrum errors and their correlations. The blue and red contours show the 68% and 95.4% confident regions, respectively.
• Figure 3: Distribution of the MCMC points for the DR12 data post-recon catalogues where the qpm covariance matrix is used, in terms of the BAO peak position variables ($\alpha_0$ and $\alpha_2$, in the top panels) and the AP parameters ($\alpha_\parallel$ and $\alpha_\perp$, in the bottom panels) for the LOWZ sample (left panels) and for the CMASS sample (right panels). In blue lines the $1\sigma$ ellipses ($\Delta\chi^2=2.30$) and in red lines the $2\sigma$ ellipses ($\Delta\chi^2=6.17$) corresponding to Gaussian fits to the likelihood based on the parameters given Table \ref{['table:data_results']}. For all cases the distribution of points is close to a Gaussian distribution.