Robust New Statistic for fitting the Baryon Acoustic Feature

TL;DR

This paper evaluates a new statistic, $oldsymbol{ ext{omega}_ ext{l}(r_c)}$, for analyzing Baryon Acoustic Oscillations, demonstrating its robustness and comparable performance to traditional methods without needing nuisance parameters.

Contribution

The study introduces and tests the $oldsymbol{ ext{omega}_ ext{l}(r_c)}$ statistic for BAO analysis, showing it performs similarly to multipole fitting but with fewer nuisance parameters.

Findings

$oldsymbol{ ext{omega}_ ext{l}(r_c)}$ matches multipole analysis in stability.

It performs well under data distortions and systematics.

No extra nuisance parameters are needed for $oldsymbol{ ext{omega}_ ext{l}(r_c)}$.

Abstract

We investigate the utility and robustness of a new statistic, $\omega_{\ell}\left(r_{c}\right)$, for analyzing Baryon Acoustic Oscillations (BAO). We apply $\omega_{\ell}\left(r_{c}\right)$, introduced in Xu et al. (2010), to mocks and data from the Sloan Digital Sky Survey (SDSS)-III Baryon Oscillation Spectroscopic Survey (BOSS) included in the SDSS Data Release Eleven (DR11). We fit the anisotropic clustering using the monopole and quadrupole of the $\omega_{\ell}\left(r_{c}\right)$ statistic in a manner similar to conventional multipole fitting methods using the correlation function as detailed in (Xu et al. 2012). To test the performance of the $\omega_{\ell}\left(r_{c}\right)$ statistic we compare our results to those obtained using the multipoles. The results are in agreement. We also conduct a brief investigation into some of the possible advantages of using the $\omega_{\ell}\left(r_{c}\right)$ statistic for BAO analysis. The $\omega_{\ell}\left(r_{c}\right)$ analysis matches the stability of the multipoles analysis in response to artificially introduced distortions in the data, without using extra nuisance parameters to improve the fit. When applied to data with systematics, the $\omega_{\ell}\left(r_{c}\right)$ statistic again matches the performance of fitting the multipoles without using nuisance parameters. In all the analyzed circumstances, we find that fitting the $\omega_{\ell}\left(r_{c}\right)$ statistic removes the requirement for extra nuisance parameters.