The U.C. Berkeley
Infrared Spatial Interferometer Array
The scientific problems that can be investigated with the ISI range from the early to the late stages of stellar evolution, and to astrometry. To date, most of the observing time has been spent on the observation of late-type stars and their dust shells, as described below. The ISI system also can be outfitted with a filterbank that allows studying the circumstellar environment not only with high spatial resolution, but also at the same time with high spectral resolution. This allows scientists to determine where exactly in the circumstellar shell organic and inorganic molecules occur, and provides clues as to how they are formed. With the recent addition of a third telescope, the ISI is now capable of measuring three baselines simultaneously, as well as a quantity known as the closure phase, which will provide us with information about the asymmetry of stellar objects.
An interferometer like the ISI is well suited for making very precise measurements of positions of stars and other celestial objects. Investigations in the field of astrometry will help to tie the astronomical reference frames together that were established in the radio, infrared and visible region of the electromagnetic spectrum, and particularly to measure positions of infrared stars which are hardly detectable in visible light. One important application could be high-precision tracking of interplanetary spacecraft equipped with infrared lasers. Another application is the precise measurement of gravitational light deflection by the Sun and the larger planets.
Whenever a new instrument with higher resolution is being built, one might eventually discover things that were completely unexpected. The Infrared Spatial Interferometer works like a new microscope on the sky that reveals much finer details than have ever seen before in the 11 micron wavelength region.
Dust Shells: [top]
Interest of the ISI scientists has mainly been focused on the study of evolved stars. These stars are further along in their life cycle compared with the Sun and typically return some fraction of their mass to the circumstellar environment. There, by cooling off, this material condenses into dust grains that absorb visible star light and re-radiate this energy in the infrared region of the spectrum. Because of this and the fact that most of the dust formation takes place within a relatively speaking small volume around those stars, high-resolution interferometry in the infrared is very well suited for studying this phase of stellar evolution.
An initial survey of 13 program stars is described in a paper by Danchi et al., 1994. Approximately half of those stars surveyed emit dust episodically, with outbursts separated by decades or centuries. Hence there are dust shells far from their photospheres. These stars include the supergiants Alpha Orionis, Alpha Scorpii, and Alpha Herculis. The dust around these stars has a mean distance from the photosphere of 38 stellar radii. Other stars observed have dust less than 6 stellar radii from their photospheres, and these include the carbon star IRC +10216, the mira variables R Leonis, Omicron Ceti, and NML Tauri, the supergiants VX Sagitarii and VY Canis Majoris, and the symbiotic star R Aquarii. The mean distance between the inner radius of the dust shells and the photospheres of these stars is 3.5 - 1.8 stellar radii. For some stars, such as Chi Cygnii and W Aquilae (both S-type stars), the distance from the star to dust changes with time, indicating that the amount of dust they throw off is variable.
The ISI is now in a program of continual monitoring of 20-30 stars, in order to observe changes in their visibilities due to variations in stellar luminosity and movements and changes in the dust surrounding these stars. Dust features have been observed to show spatial motion on time scales as short as one year, as has been described by Hale et al. 1997. By measuring dust shell expansion we are thus able to determine also stellar distances.
By moving the telescopes to very long baselines, which means telescope separations of about 56 meters, the ISI is capable of very high resolutions, approximately 20 milli-arcseconds (or 20/1000 of an arcsecond), enough to precisely resolve the diameters of many stars. The mid-infrared is believed to be a very good wavelength for measuring stellar diameters, because it is much less affected by a phenomenon known as limb darkening, which often troubles scientists working at shorter wavelengths trying to measure stellar sizes.
One of the many advantages of the ISI's narrow heterodyne detection bandwidth is that we can tune our detection wavelength to be deliberately in or out of known spectral lines. This allows us to also make accurate diameter determinations. Weiner et al., 2000 have observed larger diameters in the mid-infrared than other scientists previously measured at shorter wavelengths. By repeating measurements in a known H2O water line feature we could observe the distribution of hot water vapor in the surrounding stellar environment.
The visibility data for Betelgeuse and Antares from 32m and 56m baselines cover a sufficiently large range of spatial resolution to make a determination of their diameters possible. These diameters are 54.7±0.3 and 44.4±2 milli-arcseconds, respectively (Bester et al., 1996 and Weiner et al., 2000). They are the first diameters of stars ever measured in the mid-infrared with a two-telescope interferometer. The diameter for Betelgeuse of 54.7±0.3 milli-arcseconds, as measured in the 11 micron wavelength region with the ISI, is consistent with the measurement that was obtained in 1920 with the 20ft interferometer on the 100-inch telescope. Michelson and Pease found a diameter of 48.5 milliarcseconds, which when corrected for the theoretically expected limb darkening in visible light due to material surrounding the star, becomes about 55 milli-arcseconds. At 11 micron the effect of limb darkening is practically negligible.
The diameter of Mira has also been measured. At a particular time, it was found to have a diameter of 47.8±0.5 milliarcseconds, substantially larger than previously expected. And it varied systematically in size by ±13% as the stellar intensity varied over its well-known period of approximately 330 days.
The ISI is also well-suited for spectral line research. A spectrometer constructed by A.Betz et al. was used with one of the ISI telescopes in single-dish mode to measure infrared emission lines of stratospheric ammonia (NH3) produced by the collision of Comet Shoemaker-Levy 9 with Jupiter in July 1994. Lineshapes of three different NH3 emission lines at a wavelength 10.7 microns were measured with a resolving power of ~ 107; a detailed analysis of the temporal behavior and NH3 abundance distributions has recently been made by Fast et al.,2001.
The ISI heterodyne receivers have also been used with a specially designed spectral filterbank, described by Monnier et al., 2000. Using off-the-shelf 60 MHz radio-frequency filters, spectral resolutions of l/Dl= 27THz/60MHz = 4.5x105 are readily obtained, enough to resolve features arising from Doppler shifts as small as ~ 0.7 km/s. A bank of 32 such filters was used to measure the line profiles of CO2 in absorption in the Martian atmosphere, and NH3 in the carbon star IRC+10216. Additionally, this bank of filters may be used in conjunction with the ISI’s correlator, allowing for interferometry on spectral lines to be carried out. Using this filterbank with the interferometer, Monnier et al., 2000 were able to locate the molecular formation regions of silane (SiH4) and ammonia (NH3) around the carbon stars IRC+10216 and VY CMa.
Closure Phase: [top]
Measured interference fringes have both an amplitude and a phase which provide us with information about the geometry of the source we are observing. However, atmospheric fluctuations distort the phase information such that with only two telescopes it becomes essentially useless at high resolution. Until very recently the ISI had only two telescopes available and hence could not make accurate determinations about the symmetry, or therefore the precise shapes of the sources observed. Therefore, even though it was sometimes suspected to not be quite correct, all sources had to be modeled as being spherically symmetric. In the infrared especially, where dust shell formation is observed, it is quite possible that clumps or other asymmetries may exist.
Now, with the recent addition and activation of a third telescope, the ISI becomes an interferometer array. With such an array it becomes possible to recover most, if not all of the lost phase information distorted by the atmosphere. Even random and unknown phase shifts, when measured around a closed triangle of baselines, will cancel out leaving essentially only the intrinsic phase information contained in the source.
On UTC 2003 July 09, The U.C. Berkeley Infrared Spatial Interferometer (ISI) succeeded in recording the first measurements of three simultaneous fringes and closure phase on a stellar source at a wavelength of 11.15 microns.
The K star Alpha Bootis was observed as a point source calibrator. Other sources, Alpha Herculis and Alpha Scorpii, showed a phase closure consistent with being a symmetric source, while the observed closure phase for Omicron Ceti appears to have a slight variation.