A 10 m Test of Disk-Regulated Angular Momentum ... - CiteSeerX

From Darkness to Light ASP Conference Series, Vol. 3  108 , 2000 T. Montmerle & Ph. Andre, eds.

A 10 m Test of Disk-Regulated Angular Momentum Among Low-Mass Pre-Main-Sequence Stars Keivan G. Stassun University of Wisconsin, 475 N. Charter St., Madison, WI 53706, USA

Abstract. We summarize the results of our recent work testing the hypothesis of disk-regulated angular momentum among low-mass PMS stars. Contrary to some previous studies, we nd no connection between slow stellar rotation and the presence of near-IR signatures of disks among young (1 Myr) low-mass stars in Orion. We use 10 m photometry to test for the presence of inner-truncated disks among PMS stars lacking near-IR disk signatures; such truncated disks are a key prediction of magnetic disk-locking models. In general, we nd that stars lacking near-IR excesses do not possess disks, truncated or otherwise. Evidently, young stars can exist as slow rotators without the aid of present disk-locking, and there exist very young stars already rotating at breakup velocity whose subsequent rotational evolution will not be regulated by disks. 1. Introduction Circumstellar disks have come to be seen as dominant players in the rotational evolution of low-mass stars during the pre-main-sequence (PMS) phase. In fact, most rotational evolution models today rely chie y on magnetic disk-locking to successfully connect the rotational properties of T Tauri stars (TTS) to those of zero-age main sequence (ZAMS) stars (e.g. Bouvier et al. 1997). The principal aim of this paper is to summarize recent observations, presented in Stassun et al. (1999, 2000), that challenge the current picture of disk-regulated PMS rotational evolution. We nd no connection between stellar rotation period and near-IR diagnostics of circumstellar disks among a large sample of young PMS stars in the Orion Nebula Cluster (ONC). In addition, we do not nd compelling agreement between our observations and the predictions of disk-locking models when we consider in detail the evidence for disk material being present, and truncated near, the co-rotation radius for stars in our sample. We highlight our discovery of 1 Myr-old stars already rotating at breakup velocity|many of which do not possess disks. How do these stars deplete angular momentum prior to the main sequence? We suggest that the key to reconciling theory and observation may reside in the initial conditions: We nd that the dispersion of rotation rates among low-mass ZAMS stars is already present at 1 Myr.

1.1. Background

Stellar angular momentum is not conserved in the PMS phase (e.g. Stauer & Hartmann 1987). Traditional (i.e. Skumanich-like) wind-braking prescriptions 1

2

Stassun

have by themselves proved insucient for explaining the angular momentum evolution of PMS stars, primarily because it is dicult for winds alone to explain the large dispersion of rotation rates on the ZAMS (i.e. the presence of both slow and ultra-fast rotators; Krishnamurthi et al. 1997 and others). Implicit here are the assumed initial conditions: Previous observations have suggested that stars begin their PMS evolution with a small dispersion of rotation rates (e.g. Choi & Herbst 1996). Wind-based models have thus had diculty accounting for the apparent increase in the dispersion of stellar rotation during the PMS phase.

1.2. The current picture: Disk-regulated rotation

Magnetic disk-locking models have since come to the fore as a means for explaining the angular momentum evolution of PMS stars. Figure 1 depicts schematically the salient features of standard models (e.g. Ostriker & Shu 1995): The stellar magnetic eld \locks" the star into co-rotation with the disk at a radius (the \co-rotation radius", Rc ) that is set by the balance between the mass accretion rate, M_ , and the stellar magnetic eld strength, B . As a result, the star is prevented from spinning up as it contracts. An important prediction of the models is that magnetic pressures clear out the inner region of the disk, and that the inner disk truncation radius, Rtrunc , is approximately coincident with Rc (e.g. Armitage & Clarke 1996). Optically Thick Disk

Stellar Field Lines Co-rotation Radius

Outflows

TTS

Accretion Flows

Figure 1.

A schematic representation of magnetic disk-locking.

Edwards et al. (1993) have presented observations that they argue provide support for this picture of disk-regulated stellar rotation. Among 34 TTS in Taurus and Orion, only slowly rotating stars (Prot  8 days) in their sample showed the near-IR excess-emission signatures of circumstellar disks, while no near-IR disk signatures were observed among the more rapidly rotating TTS (Prot  2 days). In addition, the distribution of rotation periods among 130 TTS in the ONC has been reported to be strongly bimodal, suggesting the existence of two distinct rotational populations of TTS: slowly rotating disk-locked stars, and rapidly rotating disk-released stars (Herbst et al. 2000). These observations have given rise to a picture in which low-mass stars begin their PMS evolution disk-locked into slow rotation, with a relatively narrow spread of rotation periods around 8 days. Constrained by these initial conditions, rotational evolution models invoke magnetic disk-locking, and a range of disk lifetimes over which it operates (typically 1{20 Myr), to inject dispersion into the distribution of rotation rates during the PMS phase.

Testing Disk-Regulated Angular Momentum

3

2. New results We have determined rotation periods for 254 low-mass PMS stars in and around the Orion Nebula to test the hypothesis of disk-regulated rotation (Stassun et al. 1999). Unlike previous studies, which have been biased against Prot < 2 days, our study is sensitive to Prot  0:1 days. We are biased, however, against Prot > 8 days. We nd no compelling evidence that magnetic disk-locking is operating among stars in our sample, as we now discuss.

2.1. Rotation periods

The distribution of rotation periods determined by us (Fig. 2a) does not show the striking bimodality found in previous investigations of rotation in Orion; in fact, our distribution is statistically consistent with a uniform distribution.

Figure 2. (a) The distribution of Orion rotation periods determined by us is shown over the range of periods for which we are complete, 0:1 < P < 8 days. (b) The same distribution plotted as angular velocity. The arrow indicates approximately the angular velocity limit of previous studies; note our discovery of a tail of \ultra-fast" rotators. Perhaps more signi cantly, our sample includes for the rst time a signi cant population of stars with periods as short as Prot  0:5 days. These newly discovered \ultra-fast" rotators give the angular velocity distribution (Fig. 2b) a form very similar to the v sin i distribution observed among low-mass Pleiads (e.g. Queloz et al. 1998). The fact that we observe few stars with angular velocities in excess of 2 days?1 (Prot = 0:5 days), well short of our detection limit of 10 days?1 (Prot = 0:1 days), is particularly signi cant: This cuto at high angular velocities corresponds to breakup velocity for the stars in our sample (Fig. 3). A population of stars rotating at breakup is already present at 1 Myr!

Figure 3. The fraction of breakup velocity, v=vbr , is plotted vs. Prot for stars in our sample. Short-period stars rotate near breakup velocity.

4

Stassun

2.2. Circumstellar disks

Using near-IR photometry from Hillenbrand et al. (1998) for the stars in our sample, we nd no correlation between rotation period and the presence of near-IR excess-emission signatures of circumstellar disks (Fig. 4). We nd nearIR signatures of disks among stars at all rotation periods, including among the newly discovered \ultra-fast" rotators (i.e. Prot < 2 days).

Figure 4.
(I ? K ), the near-IR \excess" emission (de ned as the dierence between the observed, de-reddened, I ? K color and the photospheric color), is plotted vs. stellar Prot . Stars above the horizontal line show signi cant excess emission indicative of circumstellar disks.

2.3. Disk truncation

While the various prescriptions for magnetic star-disk coupling dier in detail, a key prediction of the models is that the stellar magnetosphere will truncate the disk at a radius, Rtrunc , that is very nearly equal to the co-rotation radius, Rc . Consequently, stars with Rc  R? (i.e. slow rotators) should have Rtrunc  R? , and so are not expected to show their disks at near-IR wavelengths (e.g. Kenyon et al. 1996). Ironically, while previous authors have upheld the disk-locking hypothesis by linking excess near-IR emission to the slow rotators, it is the slow rotators that lack near-IR excesses that may be most consistent with the models. An important question, therefore, is whether slow rotators lacking nearIR excesses indeed possess truncated disks, or if they are simply diskless. In Stassun et al. (2000) we present new and existing mid-IR (primarily 10 m) photometry for 32 low-mass PMS stars in the ONC and in Taurus-Auriga for which rotation periods have been determined and which do not show excess emission in the near-IR. Our aim is two-fold: First, we wish to determine whether slow rotators lacking near-IR excesses possess truncated disks, consistent with disk-locking models, or if these stars are simply diskless. Second, we wish to determine whether rapid rotators lacking near-IR excesses are in fact diskless, or if they possess truncated disks to which they may be coupled in a non-steadystate, which may keep them at sub-breakup rotation speeds. We do not nd evidence for truncated disks among this sample. As an example, Fig. 5 shows the spectral energy distribution for one star in the sample, JW 157. This star is a very slow rotator (Prot = 17 days) with no near-IR excess emission. The new 10 m measurement is consistent with the stellar photosphere. Any disk material around this star must be situated beyond  1 AU from the stellar surface, well beyond Rc .

Testing Disk-Regulated Angular Momentum

5

Figure 5. The spectral energy distribution (SED) for JW 157. Measurements are indicated by squares. This star has Prot = 17 days, Rc = 6R? , and no near-IR excess. The dashed line is a model photosphere, while the solid line is a model star+disk SED with Rtrunc = Rc . The dotted line has the same disk inclined 60 . The new N -band (10m) measurement clearly indicates the lack of disk material at Rc . This star, like the others for which we have obtained 10m photometry, shows no evidence for the presence of a disk. Clearly, disk-locking cannot be a factor in the subsequent angular momentum evolution of these stars. Apparently, slow rotators can exist at 1 Myr without the aid of present disk-locking. In addition, there exist very young stars already rotating near breakup velocity whose subsequent angular momentum evolution will not be regulated by disks. At a more detailed level, the new 10 m results indicate that disks, when present, do not truncate in the manner required by disk-locking models for the regulation of stellar angular momentum. The fact that young stars with disks generally possess strong near-IR excesses suggests that Rtrunc  Rc is typical of TTS. As Armitage et al. (1999) have pointed out, this could pose serious problems for models that employ magnetic coupling to disks as a means for regulating stellar angular momentum in the PMS phase.

3. Conclusions: Initial conditions revisited As we have seen, we do not nd compelling evidence in support of disk-regulated angular momentum of PMS stars. We nd no connection between stellar rotation periods and near-IR excess-emission signatures of disks among low-mass, 1 Myrold PMS stars in the ONC. Moreover, we do not nd strong agreement between our observations and the predictions of the disk-locking hypothesis when we consider the evidence for disk truncation at the co-rotation radius. A recurring theme in our discussion has been the importance of the \initial conditions" for determining what role disks play, if any, in regulating stellar rotation. From the standpoint of rotational evolution models, the main appeal of magnetic disk-locking has been its ability to connect the distribution of rotation rates observed among the PMS population|the \initial" state|to that observed among the main sequence population. Magnetic disk-locking can help explain the apparent increase in the dispersion of rotation rates during the PMS phase.

6

Stassun

But our discovery of a population of PMS \ultra-fast" rotators suggests that rotational evolution models have not assumed the correct initial conditions. We nd 1 Myr-old stars|some with disks, some without|with rotation periods spanning the range 0:5 < Prot < 10 days. In fact, when we compare the v sin i distribution of the stars in our sample (1 Myr) to that of low-mass Pleiads (100 Myr), we nd that the dispersion of rotation rates observed on the ZAMS is already present at 1 Myr. This is shown in Figure 6. Considering this, and given our observations indicating no connection between disks and PMS rotation, we pose the question: Are disks still needed to explain the rotational evolution of PMS stars? Might wind-based models now suce, given the new initial conditions? Apparently, the dispersion of rotation rates present among the main sequence population need no longer be generated during the PMS phase, but merely preserved.

Figure 6. The v sin i distribution of low-mass ONC stars is compared with the v sin i distribution of low-mass Pleiads. The dispersion of rotation rates observed on the ZAMS is already present at 1 Myr.

References Armitage, P. J., Clarke, C. J., & Tout, C. A. 1999, MNRAS, 304, 425 Armitage, P. J., & Clarke, C. J. 1996, MNRAS, 280, 458 Bouvier, J., Forestini, M., & Allain, S. 1997, A&A, 326, 1023 Bouvier, J., Cabrit, S., Fernandez, M., Martin, E., & Matthews, J. 1993, A&A, 101, 495 Choi, P. I., & Herbst, W. 1996, AJ, 111, 283 Edwards, S. E., et al. 1993, AJ, 106, 372 Hillenbrand, L. A., et al. 1998, AJ, 116, 1816 Hillenbrand, L. A. 1997, AJ, 113, 1733 Kenyon, S. J., Yi, I., & Hartmann, L. 1996, ApJ, 462, 439 Krishnamurthi, A., Pinsonneault, M., Barnes, S., So a, S. 1997, ApJ, 480, 303 Ostriker, E. C., & Shu, F. H. 1995, ApJ, 447, 813 Queloz, D., Allain, S., Mermilliod, J.-C., & Bouvier, J. 1998, A&A, 335, 183 Stassun, K. G., Mathieu, R., Vrba, F., Mazeh, T., Henden, A. 2000, AJ, in press Stassun, K. G., Mathieu, R. D., Mazeh, T., & Vrba, F. J. 1999, AJ, 117, 2941 Stauer, J. R., & Hartmann, L. W. 1987, ApJ, 318, 337