Galaxy Morphology Classification: Are Stellar Circularities Enough?

TL;DR

This study investigates whether stellar orbital circularity can serve as an effective, simple proxy for galaxy morphology classification, validated against detailed decompositions in simulations and applied to large datasets.

Contribution

It introduces a threshold-based method using stellar circularity to classify galaxy morphology, offering a computationally efficient alternative to traditional approaches.

Findings

Circularity correlates with disk and bulge components.

A threshold of 0.25 effectively distinguishes galaxy types.

Morphology-density relation aligns with observations.

Abstract

We present a preliminary study exploring whether the stellar orbital circularity of simulated galaxies, available from precomputed catalogs in the IllustrisTNG project, can be used as a proxy for broad morphological classification. We focus on the publicly available "Stellar Circularities, Angular Momenta, Axis Ratios" catalog, which enables a simple kinematic decomposition of the stellar component into disk and spheroid subsystems. By validating this approach against the more detailed five-component kinematic decomposition in TNG50, we confirm that the circularity-based disk fraction correlates most strongly with the thin disk, while the bulge fraction broadly represents the combined contribution of classical bulges and stellar halos. We then apply this decomposition to galaxies in the TNG100 simulation at redshift $z=0$ and identify a data-motivated threshold of $\mathrm{F_{disk}} = 0.25$ to distinguish early- and late-type galaxies. This threshold, lower than the commonly adopted value of $0.4$, better captures the diversity of disk-dominated systems and avoids excluding galaxies with moderately prominent disks. Additionally, we identify irregular or morphologically complex systems based on galaxies with low total disk and spheroid mass fractions. Using this classification, we recover a morphology-density relation that is broadly consistent with observations: late-type galaxies dominate in the field, while early-type galaxies are the most prevalent morphological type in clusters. Our results demonstrate that stellar circularity alone can serve as an accessible and computationally efficient morphological proxy. We also discuss the potential for this classification to support machine learning efforts as a baseline or training set for future morphological studies.