On the derivation of radial velocities of SB2 components: a "CCF vs TODCOR" comparison

TL;DR

This paper compares two methods for deriving radial velocities of SB2 binary star components, finding that TODCOR generally provides more reliable results with less zero-point shift than the traditional cross-correlation function fitting approach.

Contribution

The study systematically compares CCF and TODCOR methods for SB2 RV derivation, highlighting TODCOR's advantages in accuracy and zero-point stability.

Findings

TODCOR yields more consistent minimum masses.

Zero-point shifts are smaller with TODCOR.

Overall, TODCOR is preferable for SB2 RV analysis.

Abstract

The radial velocity (RV) of a single star is easily obtained from cross-correlation of the spectrum with a template, but the treatment of double-lined spectroscopic binaries (SB2s) is more difficult. Two different approaches were applied to a set of SB2s: the fit of the cross-correlation function with two normal distributions, and the cross-correlation with two templates, derived with the TODCOR code. It appears that the minimum masses obtained through the two methods are sometimes rather different, although their estimated uncertainties are roughly equal. Moreover, both methods induce a shift in the zero point of the secondary RVs, but it is less pronounced for TODCOR. All-in-all the comparison between the two methods is in favour of TODCOR.