Integration of the atmospheric fluctuations in a dual-field optical interferometer: The short exposure regime

Spatial phase-referencing in dual-field optical interferometry is reconsidered. Our analysis is based on the 2-sample variance of the differential phase between target and reference star. We show that averaging over time of the atmospheric effects depends on this 2-sample phase variance (Allan variance) rather than on the true variance. The proper expression for fringe smearing beyond the isoplanatic angle is derived. With simulations of atmospheric effects, based on a Paranal turbulence model, we show how the performances of a dual-field optical interferometer can be evaluated in a diagram "separation angle" versus "magnitude of faint object". In this diagram, a domain with short exposure is found to be most useful for interferometry, with about the same magnitude limits in the H and K bands. With star counts from a Galaxy model, we evaluate the sky coverage for differential astrometry and detection of exoplanets, i.e. the likelihood of faint reference stars in the vicinity of a bright target. With the 2mass survey, we evaluate sky coverage for phase-referencing, i.e. avaibility of a bright enough star for main delay tracking in the vicinity of any target direction.