JWST imaging of omega Centauri: II. Evidence for a split white dwarf cooling sequence in the near-infrared

TL;DR

The paper addresses how multiple stellar populations manifest in the white dwarf cooling sequence of omega Centauri by combining HST optical data with JWST NIR observations. Through artificial-star tests and comparisons with CO-core and He-core WD cooling tracks, the authors detect a persistent split in the WD CS extending to cooling ages of about $1$ Gyr, identifying two WD populations with masses around $0.54 M_sun$ (CO-core) and $0.46 M_sun$ (He-core). They show that the fraction of He-rich descendants decreases with radius and that the two WD groups display distinct kinematic signatures, aligning with the spatial and dynamical differences seen in the MS progenitors. This work demonstrates the viability of using optical+NIR WD CMDs to trace the imprint of mPOPs across a globular cluster’s evolution and highlights the value of JWST for studying faint WD populations in low-density outer regions.

Abstract

We present a detailed analysis of the white dwarf cooling sequence (WD CS) in omega Centauri based on combined Hubble Space Telescope (HST) and JWST observations. Our analysis confirms the previously reported split - based on HST observations in ultraviolet filters - in the upper part of the WD CS, consistent with the presence of two distinct WD populations, and extends it to a significantly fainter and cooler limit (down to ~8000 K), corresponding to cooling ages of about 1 Gyr. We used artificial star (AS) tests and cooling models to confirm that the split is evidence of two WD populations with different masses and progenitors: one sequence of canonical WDs produced by the He-normal progenitors, and one sequence of low-mass WDs originated from the cluster He-rich component. We show that the fraction of WDs from the He-rich component in the outer regions is smaller than that found in the innermost regions. We also studied the kinematics of WDs and showed that in the outer regions, the velocity distribution of WDs from He-rich progenitors is slightly radially anisotropic, while that of canonical WDs is slightly tangentially anisotropic. Both the radial variation of the fraction of WDs from the He-rich population and the difference between their velocity distribution and that of canonical WDs are consistent with spatial and kinematic differences previously found for He-rich and He-normal main-sequence (MS) stars and in general agreement with models predicting that He-rich stars form more centrally concentrated than He-normal stars.

Figures (13)

• Figure 1: Proper-motion analysis of the sources in the field. Only stars that passed the photometric quality selections and have measurable PMs are included. (a) VPD for all selected sources. (b) VPD for sources that satisfy the PM selection criterion ($\mu_{\rm R} < 3$ mas yr$^{-1}$; see panels e and i). (c) VPD for sources that do not meet the PM selection. (d) HST-based CMD for all selected sources, where the WD CS is highlighted in blue within the green fiducial lines. The 5$\sigma$ and 3$\sigma$ detection limits are marked in light and dark grey, respectively. (e) One-dimensional PM ($\mu_{\rm R}$) as a function of the HST magnitude ($m_{\rm F606W}$), with the PM selection threshold (red vertical line) separating cluster members from field stars. (f) Same CMD as (d), but only for stars that pass the PM selection. (g) Same CMD as (d), but only for stars that fail the PM selection. (h) JWST-based CMD for all selected sources. (i) One-dimensional PM ($\mu_{\rm R}$) as a function of the JWST magnitude ($m_{\rm F150W2}$). (j) Same CMD as (h), but only for stars that pass the PM selection. (k) Same CMD as (h), but only for stars that fail the PM selection. In all panels except (a), (d), and (h), cluster members (i.e., stars with $\mu_{\rm R} < 3$ mas yr$^{-1}$) are shown as dots, while non-members are marked with crosses.
• Figure 2: $m_{\rm F275W}$ versus $m_{\rm F275W} - m_{\rm F438W}$ CMD of the central field of $\omega$ Cen, based on the catalogue from 2014ApJ...797..115B. These data correspond to the HST observations that enabled the discovery by 2013ApJ...769L..32B of the split in the upper part of the WD CS, down to an effective temperature of approximately $T_{\rm eff} \sim 15{,}500$ K.
• Figure 3: $m_{\rm F606W}$ versus $m_{\rm F606W} - m_{\rm F814W}$ CMD of the stars shown in Fig. \ref{['Bellini1']}, with the same colour-coding. Photometry from 2014ApJ...797..115B. Note how the two separate groups identified in Fig. \ref{['Bellini1']} appear here as a tangled mixture. Crowding and incompleteness hinder the photometric precision required to resolve the two sequences in this optical CMD.
• Figure 4: Photometric uncertainties estimated from ASs. Panel (a) shows the results for the F150W2 filters, while panel (b) refers to the F322W2 filter. In each panel, the black points represent the difference between the recovered and input magnitudes of the ASs, plotted as a function of input magnitude. The red points mark the median values in 0.5-mag bins, and the error bars indicate the corresponding dispersions ($\sigma$) obtained from the 2.5$\sigma$-clipped distributions. The dashed blue line marks zero.
• Figure 5: Same as Fig. \ref{['err_jwst']} but for the F606W (a) and F814W (b) filters.