The effects of atmospheric calibration errors on source model parameters

Optical long-baseline interferometric data is commonly calibrated with respect to an external calibrator, which is either an unresolved source or a star with a known angular diameter. A typical observational strategy involves acquiring data in a sequence of calibrator-target pairs, where the observation of each source is obtained separately. Therefore, the atmospheric variations that have time scales shorter than the cadence between the target-calibrator pairs are not always fully removed from the data even after calibration. This results in calibrated observations of a target star that contain unknown quantities of residual atmospheric variations. We describe how Monte Carlo simulations can be used to assess quantitatively the impact of atmospheric variations on fitted model parameters, such as angular diameters of uniform-disk models representing semi- and fully-resolved single stars.