Performance and error budget of a GLAO system

Reliable prediction of the image quality delivered by a ground-layer AO (GLAO) system can be made using several methods, given the parameters of the instrument and the statistics of the turbulence profile above the observatory. Extensive data accumulated over several years at different sites indicate that GLAO can bring a significant gain in performance for classical astronomical observations, and at visible wavelengths. This gain can become spectacular on nights with calm upper atmosphere that happen infrequently but regularly at all sites studied so far. Starting from an ideal GLAO system, the influence of various instrumental errors (wave-front sensing, time delay, tilt, optics ripple, etc.) is studied, leading to the formulation of an error budget for GLAO. As the performance metrics of GLAO and classical AO are very different, we propose a new criterion for balancing error contributions.