Evolution of spatio-kinematic structures in star-forming regions: are Friends of Friends worth knowing?

TL;DR

This study evaluates the Friends of Friends algorithm's effectiveness in identifying groups in star-forming regions through N-body simulations, revealing its limitations in distinguishing different dynamical states and initial conditions.

Contribution

The paper critically assesses the Friends of Friends algorithm's ability to characterize star-forming regions, highlighting its limitations and urging caution in its application.

Findings

The algorithm identifies 1-25 groups regardless of initial conditions.

No correlation between initial group count and evolutionary tracers.

The algorithm cannot reliably distinguish between collapsing and expanding regions.

Abstract

The Friends of Friends algorithm identifies groups of objects with similar spatial and kinematic properties, and has recently been used extensively to quantify the distributions of gas and stars in young star-forming regions. We apply the Friends of Friends algorithm to $N$-body simulations of the dynamical evolution of subvirial (collapsing) and supervirial (expanding) star-forming regions. We find that the algorithm picks out a wide range of groups (1 -- 25) for statistically identical initial conditions, and cannot distinguish between subvirial and supervirial regions in that we obtain similar mode and median values for the number of groups it identifies. We find no correlation between the number of groups identified initially and either the initial or subsequent spatial and kinematic tracers of the regions' evolution, such as the amount of spatial substructure, dynamical mass segregation, or velocity dispersion. We therefore urge caution in using the Friends of Friends algorithm to quantify the initial conditions of star formation.