Accretion and the evolution of T Tauri disks

Using results and calibrations from a previous paper (Gullbring et al. 1997), we estimate disk accretion rates for pre-main-sequence stars in the Taurus and Chamaeleon I molecular cloud complexes. The median accretion rate for T Tauri stars of age similar to 1 Myr is similar to 10(-8) M. yr(-1); the intrinsic scatter at a given age may be as large as 1 order of magnitude. There is a clear decline of mass accretion rates (M) over dot with increasing age t among T Tauri stars. Representing this decline as (M) over dot proportional to t(-eta), we estimate 1.5 less than or similar to eta less than or similar to 2.8; the large uncertainty is due to the wide range of accretion rates at a given age, the limited age range of the sample, and errors in estimating stellar ages and accretion luminosities. Adopting values of eta near the low end of this range, which are more likely given probable errors and the neglect of birthline age corrections, masses accreted during the T Tauri phase are roughly consistent with disk masses estimated from millimeter-wave dust emission. Similarity solutions for evolving, expanding disks are used to investigate observational constraints on disk properties employing a minimum of parameters. For an assumed power-law form of the disk viscosity with radius nu proportional to R(gamma), eta greater than or similar to 1.5 1.5 corresponds to gamma greater than or similar to 1. The limit gamma similar to 1 corresponds to a roughly constant "alpha" in the Shakura-Sunyaev (1973) viscosity parameterization; using current observed disk sizes, we estimate alpha similar to 10(-2) (on scales similar to 10-100 AU). Much of the observed variation in mass accretion rates can be accounted for by varying initial disk masses between 0.01 and 0.2 M., but this result may be strongly affected by the presence of binary companion stars. These results emphasize the need for older samples of stars for studying disk evolution.