Older Ages for 23 Pre-Main Sequence Stars in Upper Scorpius Using Dynamical Mass-Constrained Stellar Evolutionary Models

TL;DR

This study tackles persistent age discrepancies for pre-main sequence stars by deriving mass-constrained isochronal ages for 23 K–M-type stars in Upper Scorpius using dynamical masses from Keplerian disk modeling. By interpolating five PMS model sets (including magnetic and non-magnetic variants) to Teff and L_bol and enforcing consistency with dynamical masses, the authors find that ages generally rise and model-to-model scatter decreases, with magnetic Feiden2016 models yielding a robust ~$9$–$10$ Myr age across the sample. The results align with older-age estimates from higher-mass members and eclipsing binaries, and imply longer protoplanetary disk lifetimes and extended planet-formation timescales, while also highlighting model-dependent biases across the mass range. Overall, dynamical-mass constraints significantly harmonize age estimates across PMS models and reinforce a scenario of longer disk evolution in Upper Scorpius.

Abstract

We present revised stellar ages for 23 pre-main sequence K- and M-type stars in the Upper Scorpius star-forming region, derived by using stellar dynamical masses to constrain isochronal ages from five pre-main sequence stellar evolutionary models. We find that mass-constrained stellar ages for all model sets are more consistent with the older, ~8-11 Myr age for Upper Sco derived using earlier-type stars. Additionally, applying the independent mass constraint to isochronal ages tends to 1) increase stellar ages for most model sets, and 2) decrease age scatter for individual sources between model sets. Models that account for global magnetic fields consistently produce the best match to our observations: they change comparatively little when the mass constraint is applied, and produce 9-10 Myr ages under both unconstrained and mass-constrained conditions. Most standard (nonmagnetic) models produce younger ages (3-5 Myr) when unconstrained, but older ages (6-9 Myr) when constrained by dynamical mass. Our results are consistent with literature findings that suggest median disk lifetimes may be >2x longer than previously thought.

Figures (4)

• Figure 1: Isochronal masses and ages for example source J16020757$-$2257467. Left: Contour plots of isochronal age ($\tau_{\rm all}$) versus isochronal mass ($M_{iso}$) for all five model sets considered in this work. The vertical gray bar denotes the disk-derived dynamical mass, including uncertainties. The portions of each set of contours that overlap the gray bar are what comprise our mass-constrained age ($\tau_{\rm dyn}$) for a given source. Center: Histograms of the distribution of isochronal mass for each of the five model sets. The gray bar again denotes the derived dynamical mass, including uncertainties. Right: Histograms of the distribution of unconstrained isochronal age ($\tau_{\rm all}$) for each model set. Note the logarithmic scale on the x-axis. The mass-constrained ages for this source ($\tau_{\rm dyn}$) are: BHAC15: 11$\pm$8 Myr, PARSEC v1.1: 8$\pm$5 Myr, PARSEC v1.2S: 6$\pm$4 Myr, nonmagnetic Feiden2016: 10$\pm$5 Myr, magnetic Feiden2016: 9$\pm$5 Myr.
• Figure 2: Box-and-whisker plots showing the distribution of stellar ages for each of the five sets of isochrones in this work. Median values are shown as colored lines in each box, and the left and right edge of each box are the first and third quartile values, respectively. The lower and upper whiskers end at the last data point to fall within Q1 - 1.5IQR and Q3+1.5IQR, respectively, where IQR is the inter-quartile range (Q3 - Q1). Fliers are shown as open circles. Left, Top: Distribution of ages for each model set using all ages returned by the isochrones within 3$\sigma$ (i.e. the 99.7% percentile). The median stellar age across all model sets is shown as a dashed gray line. This panel excludes sources that do not have a corresponding mass-constrained age; see Table \ref{['ages_table']} for those data. Left, Bottom: Distribution of ages for each model set using only the tracks within the 99.7% percentile that also return stellar masses that agree with the dynamical mass within uncertainties. These are referred to as "mass-constrained ages" in the text. The median mass-constrained stellar age is shown as a dashed gray line. The median age across all isochrones increases from 7$\pm$4 Myr in the unconstrained case to 9$\pm$6 Myr when the mass constraint is applied. Right Column: Histograms showing the change in stellar age ($\Delta\tau$$=$$\tau_{constrained}$ - $\tau_{\rm all}$) for each source, by model set. Bins are 2 Myr wide. $\Delta\tau$$=$ 0 is shown as a grey dotted line in all panels. The BHAC15, PARSEC v1.1, and non-magnetic Feiden2016 models tend to show an increase in age (median $\Delta\tau$$=$ (2-3) $\pm$ (3-4) Myr), while the magnetic Feiden2016 and PARSEC v1.2S model sets show a median of no change in age (median $\Delta\tau$$=$ 0 $\pm$ 3 Myr for both).
• Figure 3: Trends in stellar age ratio and mass ratio with dynamical mass, mass-constrained age, and mass ratio. From top to bottom, each row shows results for the BHAC15, PARSEC v1.1, PARSEC v1.2S, nonmagnetic, and magnetic Feiden2016 model sets. Sources with known or candidate companions are grey in all panels, but retain the symbol shape associated with each model set. From left to right, each column shows: mass ratio versus dynamical mass, mass ratio versus mass-constrained age, age ratio versus dynamical mass, age ratio versus mass-constrained age, and age ratio versus mass ratio. In all but the right-hand column, a ratio of 1:1 in the y-axis variable is indicated by a black, dashed horizontal line. In the right-hand column, the black dashed line represents a 1:1 (linear) relationship between age ratio and mass ratio. Note: for source J16143367$-$1900133, the PARSEC v1.2S models have an age ratio of $\tau_{\rm all}$/$\tau_{\rm dyn}$ = 7. For visual clarity, the y-axes of the third, fourth, and fifth columns only extend to $\tau_{\rm all}$/$\tau_{\rm dyn}$$=$ 4, so this point is not visible. Were it visible, this source would appear in the middle row at (0.23,7) in column 3, (2,7) in column 4, and (2.2,7) in column 5.
• Figure 4: Stellar age versus mass for our unconstrained (open diamonds) and mass-constrained (solid diamonds) sample using the magnetic models of Feiden2016. Typical (median) error bars for the two cases are denoted by the boxed symbols on the right-hand side of each panel. Larger samples of K- and M-type members from Upper Sco are shown as solid gray circles in both panels. Membership in Upper Sco is determined using the membership classifications of ( left) Luhman2022 and ( right) Ratzenbock2023a. Stellar masses and ages for the larger samples are calculated using the $L_{\rm bol}$ and $T_{\rm eff}$ of Fang2025 and the magnetic models of Feiden2016. There is a clear "pile-up" of sources at $\approx$0.085 M$_{\odot}$, which is the low-mass edge of the magnetic Feiden2016 model grid. We calculate KS and Anderson-Darling statistics for our sample versus both populations, and find no statistically-significant difference between our sample and the larger Upper Sco samples shown here.