John L. Hershey

Wide-angle astrometry with the Mark III stellar interferometer

Astrometric measurements made with the Mark III stellar interferometer on five nights in August-September 1988 yielded average formal 1-sigma errors for 12 FK5 stars of 6 mas in declination and 10 mas in right ascension. This improvement in precision over previously reported measurements with this instrument made in 1986 is attributable to several factors: a second 12 m baseline; oriented E-S, was added to the instrument to improve the determination of right ascension; two-color analysis was included in the data-reduction process, along with a new central-fringe identification algorithm using three spectral channels, in order to reduce atmospheric errors; thermal control was greatly improved; and changes were made to observational procedures and hardware to monitor variations in the delay offset due to residual thermal drifts. Approximately half of the new positions are within 50 mas of their FK5 positions. However, an extended series of measurements are needed to ascertain the accuracy that can be achieved by interferometry.

Application of interferometry to optical astrometry

An optical interferometer capable of tracking phase and measuring fringe visibility has demonstrated the ability to measure the precise positions of stars over large angles. This instrument has tracked phase over periods of many hours while switching sequentially among several stars. The 3.1 m separation of the siderostats has been m~asured to an accuracy of 50 ,u, indicating positional accuracies of 3 arcsec. The formal error of the least-squares solution for the baselines is of the order of a micron. The major limitations to accuracy were thermal instabilities and unmonitored siderostat positions. With improvements, this technique should be capable of astrometric accuracies exceeding one-hundredth of an arcsecond. Two small mirrors, Si and S2, intercept light from a star which is combined at beam spiitter BS and detected at photodetector P. The variable delay line DL makes interference fringes possible by adjusting for the difference between the two path lengths from the beam splitter to the stellar wave front via each arm of the interferometer. This adjustment, commonly called the delay, must change with time in order to compensate for the rotation of the Earth when tracking celestial sources, as well as to track out atmospheric turbulence. The fringe pattern, i.e., the intensity of the light at the photodetector as a function of delay, is proportional to the Fourier transform of the optical bandpass of the system. The sensitivity of the interferometer is proportional to the