On the determination of the He abundance distribution in globular clusters from the width of the main sequence

TL;DR

This paper demonstrates that the (F606W-F814W) colour width of the main sequence in globular clusters is a reliable and robust indicator of the initial helium abundance variation among sub-populations, aiding the study of multiple stellar populations.

Contribution

It shows that the (F606W-F814W) colour difference is an accurate and model-insensitive method to estimate helium abundance spreads in globular clusters using UV/optical data.

Findings

The (F606W-F814W) colour width reliably indicates helium abundance variations.

The derivative dY/d(F606W-F814W) is insensitive to stellar model assumptions.

The method is robust against physical assumptions in stellar models.

Abstract

One crucial piece of information to study the origin of multiple stellar populations in globular clusters, is the range of initial helium abundances $\Delta{Y}$ amongst the sub-populations hosted by each cluster. These estimates are commonly obtained by measuring the width in colour of the unevolved main sequence in an optical colour-magnitude-diagram. The measured colour spread is then compared with predictions from theoretical stellar isochrones with varying initial He abundances, to determine $\Delta{Y}$. The availability of UV/optical magnitudes thanks to the {\sl HST UV Legacy Survey of Galactic GCs} project, will allow the homogeneous determination of $\Delta{Y}$ for a large Galactic globular cluster sample. From a theoretical point of view, accurate UV CMDs can efficiently disentangle the various sub-populations, and main sequence colour differences in the ACS $F606W-(F606W-F814W)$ diagram allow an estimate of $\Delta{Y}$. We demonstrate that from a theoretical perspective the ($F606W-F814W$) colour is an extremely reliable He-abundance indicator. The derivative d$Y$/d($F606W-F814W$), computed at a fixed luminosity along the unevolved main sequence, is largely insensitive to the physical assumptions made in stellar model computations, being more sensitive to the choice of the bolometric correction scale, and is only slightly dependent on the adopted set of stellar models. From a theoretical point of view the ($F606W-F814W$) colour width of the cluster main sequence is therefore a robust diagnostic of the $\Delta{Y}$ range.