RUBIES Survey: Unresolved Binary Stars

• RUBIES Survey is an observational program that uses Rubin Observatory’s deep photometry to identify unresolved main sequence binaries in stellar clusters.
• It employs detailed CMD analysis and Monte Carlo simulations to derive binary fractions and assess radial trends from cluster cores to halos.
• The survey refines our understanding of binary evolution and cluster dynamics by comparing outer-field and core observations, informing future simulations.

The RUBIES Survey, formally known as the Rubin Unresolved Binary Investigation in Extra‐core Structures, is an observational program leveraging the Rubin Observatory's wide-field, high-precision photometric capabilities to systematically chart the population of unresolved main sequence binary stars in stellar clusters across a broad dynamical range. Its first results, based on Data Preview 1 (DP1) of the 47 Tucanae globular cluster, reveal binary populations well beyond the cluster core and provide new empirical constraints on stellar dynamics and binary evolution mechanisms (Cordoni et al., 4 Sep 2025).

1. Scientific Objectives and Rationale

RUBIES is designed to quantify the fraction and spatial distribution of faint, unresolved main-sequence binaries in diverse environments, from dense cluster cores—subject to frequent dynamical encounters—to the sparsely populated halos, where two-body relaxation timescales approach a Hubble time. The survey pursues three major scientific aims:

• Trace the primordial binary fraction as a function of location within clusters, probing how initial binary-formation channels and subsequent dynamical processes (three-body interactions, mass segregation, core collapse) sculpt cluster-wide binary demographics.
• Constrain theoretical models of binary disruption, exchange, and survival by measuring radial trends in binary fraction out to—and beyond—the half-light radius.
• Supply empirical distributions to inform N-body and Monte Carlo simulations of cluster evolution, treating binaries as vital repositories of binding energy that affect long-term structure and dynamics.

These goals require photometric precision, spatial coverage, and robust source characterization well beyond the capabilities of previous surveys (e.g., central HST fields), making the Rubin Observatory uniquely suitable for this purpose.

2. Survey Footprint, Observational Data, and Sample Definition

The DP1 deployment includes ≈1,700 LSSTComCam exposures in ugrizy covering 47 Tucanae. For binary detection, the key dataset is the deep coadds in g, r, and i, spanning a circular annulus from 18′ to 40′ from the cluster center (∼6–12 times the half-light radius $R_h$), approaching the tidal radius (∼40′). This "outer-field" configuration is complementary to legacy HST "inner-field" studies.

Photometric properties:

• Coadd depths: $i\sim22$, corresponding to $S/N\sim50$ at $i\approx20$
• Photometric uncertainties: $\sigma_i\lesssim0.02$ mag for $i<19$, rising to $\sigma_i\lesssim0.05$ at $i\sim20.5$

Source catalogs are constructed using LSST Science Pipelines (PSTN-019), ensuring internally consistent photometric calibration and precise coadd construction.

Selection criteria for cluster membership and sample purity:

• Point-source morphology ($\text{refExtendedness}=0$)
• PSF model agreement in g, r, i bands
• Blendedness $b<0.05$ in each band to exclude blends
• High-confidence Gaia DR3 association ($p>0.8$ as per Vasiliev 2021)
• Proximity ($|\Delta(g-i)|<0.4$ mag) to a PARSEC isochrone ([M/H] = –0.5, $d=4.6$ kpc, $E(B–V)=0.025$)

Gaia completeness limitations restrict unbiased sampling to $i\sim20.5$ over the 18′–40′ annulus.

3. CMD Construction and Binary Identification Methodologies

Binary detection is performed via analysis of the $i$ vs. $g-i$ color-magnitude diagram (CMD) for high-confidence cluster members with $i=18$–$20$.

Defining mass ratio $q = M_\mathrm{sec}/M_\mathrm{prim}$, the methodology proceeds as:

• Compute a smoothed main sequence (MS) fiducial color $(g-i)_\mathrm{fid}(i)$ by median filtering in 0.5 mag bins, with dispersion $\sigma_\mathrm{fid}(i)$.
• Single-star region ("Region A"): $$[(g-i)_\mathrm{fid}(i) - 3\sigma_\mathrm{fid}(i),\, (g-i)_{q=0.7}(i)]$$, with $(g-i)_{q=0.7}(i)$ from PARSEC.
• Binary region ("Region B"): $$[(g-i)_{q=0.7}(i),\, (g-i)_{q=1}(i) + 3\sigma_\mathrm{fid}(i)]$$
  • Equal-mass binaries ($q=1$) lie $\Delta i\approx0.75$ mag above the MS.

Counts of single-star candidates ($N_A$) and high-$q$ binaries ($N_B$) yield the raw binary fraction:

$f_\mathrm{bin}(q>0.7) = \frac{N_B}{N_A + N_B}$

Field contamination is corrected using $10^5$ Monte Carlo trials with synthetic Besançon-model field stars inserted in the CMD. This yields a contamination-corrected binary fraction.

4. Key Empirical Results: Radial Trends and Comparisons

Applying the above methodology to DP1 results in $N_A + N_B = 1308$, $N_B = 25$. After contamination correction:

$f_\mathrm{bin}(q > 0.7) = 0.016 \pm 0.005$

This outer-field value (6–12 $R_h$) is notably higher than HST central field results: $f_\mathrm{bin}(q > 0.7)=0.005 \pm 0.003$ for $R < R_h$ (Milone 2012), a $1.9\sigma$ excess, indicating an environmental gradient in the binary fraction.

Assuming a flat $q$ distribution, scaling gives total binary fraction:

$f_\mathrm{tot} \simeq 0.053 \pm 0.017$

This is consistent with other core-region spectroscopic studies (e.g., MUSE: Müller-Horn 2025; Ji 2015).

5. Data Analysis Protocols and Quality Assurance

The LSST Science Pipelines deliver:

• Instrumental calibration and photometric zeropointing
• Iterative source extraction and deblending (Bosch et al. 2018)
• Multi-epoch image coaddition
• PSF fitting for robust photometry

No forced photometry on individual visits is performed, maintaining coadd-level purity. A secondary blending test counts neighbor sources within $2''$–$3''$ annuli; $<0.5\%$ of Region B sources are rejected for blending, confirming minimal contamination.

6. Theoretical Implications and Future Directions

The DP1 results extend binary studies to cluster halos, indicating Rubin's ability to resolve binary distributions beyond the core and test dynamical evolution models in environments where relaxation times rival cosmic age.

Forthcoming full LSST releases will enable:

• CMD extension to $i \sim 24.5$, probing binaries with lower primary masses and smaller mass ratios
• Forced photometry on single exposures, facilitating completeness studies and multi-epoch variability analysis
• Enhanced proper-motion membership identification, with Gaia-like accuracy out to the tidal radii
• Artificial star tests for completeness maps

These will allow RUBIES to systematically chart both primordial and dynamically processed binary fractions across hundreds of clusters throughout the Milky Way and Local Group.

7. Survey Legacy and Broader Impact

RUBIES represents a transformative advance in unresolved binary census methodologies by exploiting Rubin's large field of view, photometric depth, and precise source characterization. The demonstration in 47 Tucanae's halo validates survey strategies for studying binary evolution and cluster dynamics at previously inaccessible spatial scales, establishing empirical benchmarks for future dynamic and synthetic models (Cordoni et al., 4 Sep 2025).