Nonlinear Color-Metallicity Relations of Globular Clusters. IV. Testing the Nonlinearity Scenario for Color Bimodality via HST/WFC3 u-band Photometry of M84 (NGC 4374)

TL;DR

This study tests whether the observed bimodal color distributions of globular clusters are caused by nonlinear color-metallicity relations, using HST u-band photometry of M84 to support the nonlinear-CMR scenario.

Contribution

It provides empirical evidence supporting the nonlinear-CMR hypothesis as an explanation for globular cluster color bimodality using new u-band data.

Findings

u-z and u-g color distributions differ systematically from g-z.

Results are consistent with nonlinear-CMR model predictions.

Supports nonlinear-CMR scenario as explanation for color bimodality.

Abstract

Color distributions of globular clusters (GCs) in most massive galaxies are bimodal. Assuming linear color-to-metallicity conversions, bimodality is viewed as the presence of merely two GC subsystems with distinct metallicities, which serves as a critical backbone of various galaxy formation theories. Recent studies, however, revealed that the color-metallicity relations (CMRs) often used to derive GC metallicities (e.g., CMRs of g-z, V-I and C-T1) are in fact inflected. Such inflection can create bimodal color distributions if the underlying GC metallicity spread is simply broad as expected from the hierarchical merging paradigm of galaxy formation. In order to test the nonlinear-CMR scenario for GC color bimodality, the u-band photometry is proposed because the u-related CMRs (e.g., CMRs of u-g and u-z) are theoretically predicted to be least inflected and most distinctive among commonly used optical CMRs. Here, we present Hubble Space Telescope (HST)/WFC3 F336W (u-band) photometry of the GC system in M84, a giant elliptical in the Virgo galaxy cluster. Combining the u data with the existing HST ACS/WFC g and z data, we find that the u-z and u-g color distributions are different from the g-z distribution in a very systematic manner and remarkably consistent with our model predictions based on the nonlinear-CMR hypothesis. The results lend further confidence to validity of the nonlinear-CMR scenario as an explanation for GC color bimodality. There are some GC systems showing bimodal spectroscopic metallicity, and in such systems the inflected CMRs often create stronger bimodality in the color domain.