A Spectroscopic Technique for Measuring Stellar Properties of Pre-Main-Sequence Stars

We describe a technique for deriving effective temperatures, surface gravities, rotation velocities, and radial velocities from high-resolution near-IR spectra. The technique matches the observed near-IR spectra to spectra synthesized from model atmospheres. Our analysis is geared toward characterizing heavily reddened pre–main-sequence stars, but the technique also has potential applications in characterizing main-sequence and post–main-sequence stars when these lie behind thick clouds of interstellar dust. For the pre–main-sequence stars, we use the same matching process to measure the amount of excess near-IR emission (which may arise in the protostellar disks) in addition to the other stellar parameters. The information derived from high-resolution spectra comes from line shapes and the relative line strengths of closely spaced lines. The values for the stellar parameters we derive are therefore independent of those derived from low-resolution spectroscopy and photometry. The new method offers the promise of improved accuracy in placing young stellar objects on evolutionary model tracks. Tests with an artificial noisy spectrum with typical stellar parameters and a signal-to-noise ratio of 50 indicate a 1 σ error of 100 K in Teff, 2 km s-1 in v sin i, and 0.13 in continuum veiling for an input veiling of 1. If the flux ratio between the sum of the Na, Sc, and Si lines at 2.2 μm and the (2–0) 12CO band head at 2.3 μm is known to an accuracy of 10%, the errors in our best-fit value for log g will be Δ log g = 0.1–0.2. We discuss the possible systematic effects on our determination of the stellar parameters and evaluate the accuracy of the results derivable from high-resolution spectra. In the context of this evaluation, we quantitatively explore the degeneracy between temperature and gravity that has bedeviled efforts to type young stellar objects using low-resolution spectra. The analysis of high-resolution near-IR spectra of MK standards shows that the technique yields very accurate values for the effective temperature. The greatest uncertainty in comparing our results with optical spectral typing of MK standards is in the spectral type–to–effective temperature conversion for the standards themselves. Even including this uncertainty, the 1 σ difference between the optical and infrared temperatures for dwarfs at 3000–5800 K is only 140 K. In a companion paper, we present an analysis of heavily extincted young stellar objects in the ρ Ophiuchi molecular cloud.