Space Sciences Laboratory, University of California, Berkeley, CA 94720-7450 Introduction The Berkeley Infrared Spatial Interferometer (ISI) is located at Mount Wilson Observatory near Pasadena, CA, a site known for good “seeing” conditions. The ISI saw first light in 1988. It is presently a 3 telescope array, operating in mid- infrared (~10 micron). The 1.6 m diameter telescopes are mounted in semi-trailers, which allows them to be moved into various configurations according to observational requirements. The ISI has been used to study many characteristics of stars, including their diameters, the distribution and composition of material around them, and how they change with time. In order to study these properties, extremely high angular resolution is required since stars and the material surrounding them are so far away that they appear to be nearly point sources to most telescopes. Stellar Modeling and Closure Phase A single plot of visibility versus spatial frequency is only a one dimensional representation of a stellar object. Many visibility measurements over a series of baseline orientations must be combined in order to build up a two dimensional image. Until recently, ISI stellar modeling assumed the imaged stars and their surrounding dust were symmetric. This is because the Fourier phase information required to detect asymmetries was corrupted by atmospheric fluctuations, which introduce random optical path length changes between the telescopes and the star. With the addition of a third telescope in 2003, however, useful information about asymmetrical structure can be determined by the use of "closure phase." Atmospheric effects are cancelled by summing the phases of each of the 3 fringes from each of the 3 baselines ( "closure" coming from the need to use a closed triangle of telescopes to derive the summed phase). Stellar Evolution and The Interstellar Medium The Interstellar Medium (ISM) not only contributes material at the birth and early development of stars and planets, but also receives material throughout the evolution and late stages of their existence. The ISI is especially well suited to study late-type stars since they are typically bright in the mid-infrared, and high-resolution measurements are needed to study the material being cast off by these stars into the ISM. Most of these stars are on the so-called Asymptotic Giant Branch (AGB) in their life-cycles--as will be our Sun in a few billion years. Much can be learned on shorter time scales, however, as seen by ISI observations that show significant changes in dust shell structure and position over just a few years. The circumstellar out-flow that creates such dynamic dust shells is not well understood, nor is the history and evolution of mass loss from AGB stars. Interferometry Interferometry is an important observational technique in which light from multiple telescopes is combined to give much higher resolution than can be obtained using any one of the telescopes alone. The resolution of a telescope is determined by its diameter and operating wavelength; for an interferometer, however, the separation ("baseline") between the telescopes determines the resolution. Thus, with its maximum 75 m baselines, the ISI has a resolution comparable to a single telescope that is ~75 m in diameter, larger than any conventional telescope. Furthermore, the atmosphere is not uniform over the entire beam of large diameter telescopes so their resolution is limited by atmospheric “seeing”. Stellar interferometry was first developed successfully by Albert Michelson in the 1920's, coincidentally also at Mt. Wilson. Heterodyne Detection Unlike all other existing optical/IR interferometers, the ISI employs a very novel, indirect method of interferometry: heterodyne detection, followed by radio frequency (RF) correlation. The heterodyne detection uses a carbon-dioxide laser as an IR local oscillator (LO) to mix with the incoming IR starlight in each telescope, producing an RF signal with phase and amplitude corresponding to those of the original starlight. This RF signal can then be transported by coaxial cables rather than elaborate systems of mirrors and light pipes and can be amplified, filtered, measured, etc. by relatively common RF components. Since correlation requires the relative phase between the telescopes to be preserved, each of the 3 telescope LOs is phase-locked to a master laser. Also, each of the RF signals is delayed by the proper amount to make the travel time from the star to a central correlator equal, even as the star moves across the sky. Visibility Measurements Much like a double-slit interferometer generating patterns of visible light and dark "fringes", the ISI correlator produces electrical fringes that can be recorded and analyzed. The term "visibility" is used to describe the contrast in intensity between the peaks and troughs of the fringe pattern, measured as I max and I min respectively. The visibility of a stellar object depends on its size and shape, its position in the sky during the observation, and the length and orientation of the baseline. Maximum visibility (= 1) occurs when the object is completely unresolved, such as a point-source star at a relatively short baseline. In this case, light enters each telescope pair as a plane wave across the baseline, thus producing RF signals that match well in phase when brought together and correlated. Larger, more extended sources can be thought of as a number of point sources whose separate plane waves are not in phase and thus produce RF signals that correlate less well. Minimum visibility (= 0) occurs for longer baselines when the RF signals are 180 degrees out of phase and do not correlate at all. Many details of source size and structure can be determined from plots of visibility versus spatial frequency (which are equivalent to plotting Fourier components of the image). The Berkeley Infrared Spatial Interferometer A Heterodyne Stellar Interferometer For The Mid-Infrared R.L. Griffith, A.A. Chandler, K. Tatebe, D.D.S. Hale, E.H. Wishnow, W. Fitelson, C.H. Townes ISI 4 – 8 – 12 meter linear-baseline configuration. Laser and starlight path with correlation A spherically symmetric model of a star and surrounding dust. Radiation intensities are computed assuming radiative transfer. Asymmetrical model NML Tauri Master laser oscillator optics table Visibility measurements for NML Tauri Telescope trailer design ( Pfund design) Visibility formula Alternative Visibility formula Future Goals and Directions The 3 telescopes of the ISI were moved in 2006 from a linear, short baseline configuration to a triangular one of about 35 m sides. This layout allows visibility and closure phase measurements to be taken at an optimal resolution for studying AGB star diameters and their asymmetries. The complexities of dust shells might be partially explained by these asymmetries, which may result from hot spots on the star surface or from physical deformations of the photosphere. While the precise dynamics responsible for these asymmetries are not yet known, the ISI hopes to continue to add to our understanding. 4-8-12 m Baseline 2003 34-35-39 m Baseline 2006 Baseline configurations Telescope 1 Telescope 2 Closure phase NML Tau Telescope 3