Since the pioneering work of Connes et al. (1969), many infrared spectra of the planets have been obtained by various researchers using Fourier transform spectrometers. However, the astronomical use of FTSes in the millimeter and submillimeter wavebands has been restricted due to the lack of large, high surface accuracy telescopes sited at dry locations. The recent construction of the James Clerk Maxwell Telescope [JCMT] and the Caltech Submillimeter Observatory [CSO] (Woody et al. 1994), both located at an altitude of 13,200 feet above sea level near the summit of Mauna Kea (on the island of Hawaii), has removed this impediment, so sub-THz Fourier transform spectroscopy of astronomical sources is now possible.
In order to study the millimeter and submillimeter spectra of the giant planets at moderate resolution,
Serabyn and Weisstein (1996)
constructed an intermediate resolution (
)
Fourier transform spectrometer. The
screw-driven translation stage on which the moving mirror is mounted provides a maximum total travel of 50 cm. In the
default position of the translation stage, a one-sided travel of 46 cm gives a maximum (unapodized) spectral resolution
of 200 MHz, while a two-sided travel of 5 cm gives a maximum two-sided resolution of 3.6 GHz. (The one-sided travel plus
half the two-sided travel does not quite sum to 50 cm because a small amount of space is left at the ends for limit
switches in order to avoid the unpleasant consequences of driving the stage beyond the end of its intended travel; see
Fig. -3.) The FTS is an upgraded version of an earlier interferometer which was
used exclusively for holographic dish surface accuracy measurements (Serabyn et al. 1991). During observations, the FTS is
mounted at the Cassegrain focus of the 10.4 m CSO telescope. A view of the CSO telescope with its shutter open and dish
revealed is shown in Fig. .
Fourier transform spectrometers (e.g., Connes 1961, Vanasse and Sakai 1967, Schnopper and Thompson 1974, Oepts 1976, Brault 1985) operate by splitting an incoming beam of radiation into two parts, applying a differential phase shift, then recombining the beams. The phase shift is then varied, and an interference pattern known as an interferogram is swept out as a function of phase shift. The phase shift is most commonly produced by translating a reflecting element along one of the beams, thus producing a varying optical path difference. Because the CSO FTS is also used for holographic dish measurements, it is necessary to be able to steer the beam in one of the interferometer arms in azimuth and elevation. This consideration requires that flat surfaces be used for the end-mirrors, which in turn demands use of a dielectric beamsplitter. We therefore built a simple Michelson-type interferometer as opposed to a more complicated configuration such as a Martin-Puplett interferometer (Martin 1986) which uses right-angle corner reflectors. While the single input and output port Michelson-type interferometer has a lower sensitivity than a dual-port Martin-Puplett interferometer, its simpler optical configuration is a considerable advantage in allowing rapid alignment and setup at the sometimes mind-numbing altitude of the summit of Mauna Kea.
As shown in Fig. , our interferometer is decoupled from the F/12.4 beam of the CSO's secondary mirror using two flat mirrors (M1 and M2) placed on either side of the Cassegrain focus. This arrangement folds the beam, allowing it to expand to the necessary size despite space constraints on the underside of the telescope. After reflecting off M2, the radiation is collimated by means of an off-axis paraboloidal mirror (P1), then split into two separate beams using a dielectric mylar beamsplitter (BS). One beam is then reflected off a mirror that is translatable along its normal (I1), and the other reflects off a stationary mirror positioned at right angles to the first (I2). The reflected beams are then recombined at the beamsplitter. Finally, the superposed beams are focused onto the CSO facility bolometer by a second off-axis paraboloid (P2). The instrument is aligned by mounting a commercial HeNe laser in place of the bolometer and iteratively adjusting mirrors in order to bring all laser reflections into coincidence. Because FTSes can be operated without cryogenic cooling, the instrument is used at ambient temperature during observations (only the detector must be cooled).
Fig. shows an annotated image of the FTS mounted at the CSO Cassegrain focus. The FTS instrument is run using FTSRUN, a flexible and (for the most part) user-friendly program which runs on the CSO summit VAXstation. This program allows the near-autonomous collection of data and on-line examination of raw interferograms using a set of predefined observing commands.
 |
