Stellar Rotation in the First Six Million Years: Rotational Velocities and Radii Estimates of T Tauri Stars in IC 5070 and IC 348

TL;DR

We measure $v \sin i$ for 54 stars in IC 5070 and 99 stars in IC 348 with WIYN/Hydra, integrating with literature on temperatures, luminosities, periods, disks, and binarity to study angular-momentum evolution. The results support disk-locking effects, showing slower rotation for disk-bearing (Class II) stars than disk-free (Class III) stars in a combined IC 5070/ONC sample, while IC 348 exhibits a high fraction of slow-rotators among Class III that may reflect recent disk loss. A maximum-likelihood radii analysis, using Baraffe2015 and Somers2020 starspot models, indicates ages of about 1–2 Myr for IC 5070 and 4–6 Myr for IC 348, with substantial starspot coverage (roughly 50–85%) necessary to reconcile observed radii with models. These findings imply that starspots play a key role in radius inflation for very young PMS stars and that disk and binary status modulate rotational evolution, with implications for early stellar angular-momentum evolution and age dating of young clusters.

Abstract

We have acquired high-resolution optical spectroscopy for a sample of T Tauri stars (TTSs) in open clusters using Hydra on the WIYN 3.5m telescope, and present projected rotational velocities (v sin i values) for 54 stars in IC 5070 and 99 stars in IC 348. We combine these with published values for stellar temperature, luminosity, rotation period, circumstellar disk status, and binarity; we are predominantly interested in how the last two factors may affect the rotation speeds of the stars. We find evidence to support theories that interaction with circumstellar disks may slow the rotation of TTSs compared to Class III stars in both clusters. We also identify a higher fraction of slow-rotating Class III stars in IC 348 compared to other clusters; we suggest that some fraction of these may be stars that recently lost their disks. We find that a higher fraction of binary stars are rapid rotators compared to single stars, though not to a statistically significant degree. We also combine our v sin i measurements with rotation periods to estimate projected stellar radii, which we compare to predictions from stellar evolution models using a maximum likelihood method. We continue to show that models with increasing starspot coverage reduce radius inflation and align better with published age estimates than models without starspots.

Figures (10)

• Figure 1: Plot of our measured $v$ sin $i$ values vs. values for the same stars from Mermilliod2009, to demonstrate the effectiveness of our measurement methods. The magenta dashed line depicts a perfect 1-to-1 correlation. For objects with $v$ sin $i$$\geq$ 11 $\mathrm{km}\,\mathrm{s}^{-1}$, we find strong agreement between the measurements, excluding vB 36 as discussed in the text.
• Figure 2: Comparison of our $v$ sin $i$ measurements with measurements from Kounkel2019 (Kounkel2019; left), Cottaar2014 (Cottaar2014; center), and Nordhagen2006 (Nordhagen2006; right). Objects with $v$ sin $i$$<$ 11 $\mathrm{km}\,\mathrm{s}^{-1}$ are marked as upper limits with arrows. Objects we identified as "likely" binary systems are marked in yellow, while "possible" binaries are in pink, and "single" stars are in blue. The magenta dashed line depicts a perfect 1-to-1 correlation. The measurements from Kounkel2019 and Cottaar2014, which are both from APOGEE, tend to be higher than ours, although we generally find good agreement with Nordhagen2006.
• Figure 3: Normalized $v$ sin $i$ distributions for Resolution-Limited Sample Class II stars (blue) and Class III stars (yellow) in IC 5070 (left) and IC 5070 combined with the ONC (right). The gray bar marks the area below the velocity resolution limit. The histogram bins are 2 $\mathrm{km}\,\mathrm{s}^{-1}$ wide. The Class III stars appear to rotate faster, on average, than the Class II stars, and very few Class II stars rotate faster than 50 $\mathrm{km}\,\mathrm{s}^{-1}$. In a comparison of the distributions on the right, the K-S $p$-value is 0.0043, which is a statistically significant difference.
• Figure 4: Normalized $v$ sin $i$ distributions for Resolution-Limited Sample Class II stars (blue) and Class III stars (yellow) in IC 348. The gray bar marks the area below the velocity resolution limit. The K-S test $p$-value = 0.091, which is statistically significant at the 10% level. None of the Class II stars have $v$ sin $i$$>$ 27 $\mathrm{km}\,\mathrm{s}^{-1}$ while four of the Class III stars do, though our sample size is small.
• Figure 5: Normalized $v$ sin $i$ distributions for stars we identified as "quality single stars" (blue) and "quality likely binary" systems (yellow) in IC 5070 (left) and IC 5070 combined with the ONC (right). Only objects with $R_{TD}$$\geq$ 7 are included. The gray bar marks the area below the velocity resolution limit. In both cases, a K-S test does not yield a statistically significant $p$-value.