Telluric Line Effect on High Precision Radial Velocity Survey of K and M Dwarfs
Abstract
The red and NIR region, where K and M dwarfs emit most of light, is the desirable region for radial velocity (RV) measurements for detecting low mass planets, but this wavelength region is heavily contaminated with telluric absorption lines. Variation in the telluric line depths and centroids can result in large RV measurement uncertainties, limiting the sensitivity to detect low mass planets. Here we use simulations to study effect of telluric removal and the residuals on RV measurements and determine the level of correction needed to minimize the effect. Simulated spectra from three representative spectrographs with spectral resolutions, R=60K, 80K, 100K and 120K for wavelength coverage at 0.38-0.62 μm (called the optical spectrograph), 0.38-0.90 μm (called the broad optical spectrograph) and 0.90-2.4 μm (called the NIR spectrograph), have been studied. Two methods are used to study the RV effect by the telluric lines. The first one is a 'Masking' method, in which the telluric lines are identified and removed from RV calculation. The other method is a 'Removal' method, in which all heavily saturated lines are masked out and the remaining lines are subtracted by synthetic atmospheric spectra to a desired level. Our results show that, in case of late M dwarfs, the broad optical spectrograph can gain additional RV sensitivity over the optical spectrograph if telluric lines can be modeled and subtracted to better than 10%, or all lines deeper than 5% are masked out from RV calculation. For the earlier type stars, it requires better than 2% modeling and subtracting precision with the broad optical spectrograph to gain additional Doppler sensitivity over the optical spectrograph. Besides the photon gain with the NIR spectrograph over the optical spectrograph for late M dwarf observations, the NIR can gain additional advantage of Doppler sensitivity over the optical tool for late M dwarfs when telluric residuals can be subtracted to below 1%. However, it is never better than the broad optical spectrograph in any cases with the same residual levels. These indicate the broad optical spectrograph can potentially be an optimal spectrograph for high precision RV surveys for low mass planets.
- Publication:
-
American Astronomical Society Meeting Abstracts #227
- Pub Date:
- January 2016
- Bibcode:
- 2016AAS...22713719S