Optical interferometry is no longer on the fringe of astronomy.
Overlooking Los Angeles, six small domes nestle amid the pine trees atop Mount Wilson. Individually, the 1-metre telescopes inside those buildings have no chance of competing with the biggest ground and space telescopes. But collectively, the Mount Wilson telescopes are producing some of the sharpest images ever made.
Spread in a Y-shaped array across the top of the mountain, the telescopes are part of the Center for High Angular Resolution Astronomy (CHARA). The light from each one is funnelled through vacuum tubes to a central shed, where it is combined in a process called interferometry. Merging the light beams from the widely separated domes gives CHARA a resolving power, or sharpness, equivalent to a single telescope with a 330-metre mirror. That's more than 50 times better than the Hubble Space Telescope's resolution, allowing CHARA to see details on the surfaces of stars where other telescopes just see blurry blobs of light.
Radio astronomers have relied on interferometry for more than half a century, but optical astronomers have lagged behind. Now, optical interferometry has come of age. Several observatories are producing strong scientific results, including one reported on page 870 by researchers using the CHARA array1. At the end of last year, the team imaged a disk of dust almost as wide as the Solar System as it crept in front of a large, old star and blotted out its light. This was the first direct image of an eclipsing binary system that has puzzled astronomers for more than a century. "This is moving us into a realm that radio astronomy has been able to enjoy for decades," says Robert Stencel, an astronomer at the University of Denver in Colorado and a co-author on the paper.
This is moving us into a realm that radio astronomy has been able to enjoy for decades. ,
From the start, radio astronomers have enjoyed several advantages over optical observers. Earth's atmosphere doesn't blur radio waves as it does the shorter wavelengths of light. Moreover, radio signals gathered at separate dishes can be digitized, transmitted electronically, then recombined into an interference pattern — the basis of a high-resolution image. This ease of handling has allowed radio astronomers to amalgamate data from dishes all over the globe, creating virtual arrays with baselines as wide as Earth itself.
But with optical interferometry, astronomers must intertwine the faint light beams in real time by routing them through tunnels with nanometre-level precision. They also have to counteract the effects of atmospheric blurring using a complex technology called adaptive optics. And because many optical arrays use relatively small telescopes, they have trouble gathering enough light to study anything but bright stars nearby.
Even with those constraints, optical interferometry has yielded new insights about stars, such as how binary systems swap mass and how stars bulge when they spin. Now, astronomers are pushing the technique by combining light from more than two telescopes. Multiple beams not only make data collection more efficient — more photons are caught and used — they also provide cross-checks on the data, making it easier to build up an image from the interference pattern. CHARA first demonstrated2 a four-beam combiner in 2007, and next year it plans to try for a record six beams at once.
The advances are turning once-difficult experiments into more routine operations, opening up the process to astronomers who are not experts in interferometry. "Now we're getting more general users," says Françoise Delplancke, head of the interferometry group for the European Southern Observatory's Very Large Telescope Interferometer (VLTI) in Chile. The number of science papers based on optical interferometry has surged as well, from 9 in 1999 to 56 last year. The VLTI, the focus of European support, is responsible for about half of those.
The support for US facilities is more fragmented. CHARA is a university-run operation supported by the National Science Foundation. A potential rival, the Magdalena Ridge Observatory in New Mexico, has run into delays because of funding problems. A NASA-supported interferometer involving the twin 10-metre Keck telescopes in Hawaii was supposed to achieve VLTI-like capabilities with the addition of four to six small 'outrigger' telescopes. But the auxiliary project was derailed in 2006 over environmental and cultural concerns about building new telescopes on the Mauna Kea summit.
Yet not everyone has given up on Mauna Kea, which holds the largest concentration of huge telescopes on Earth. Guy Perrin, an astronomer at the Paris Observatory and principal investigator for the Optical Hawaiian Array for Nano-radian Astronomy (OHANA), is connecting the seven large telescopes at the summit into an array with a baseline of 800 metres. As a proof of principle, Perrin has already combined light from the two Keck telescopes via inconspicuous optical fibres, which would obviate the need to connect the telescopes with tunnels3.
Reached by telephone atop the summit, Perrin last week was busy implementing a second stage — a fibre-optic link to connect the Gemini North telescope to the Canada-France-Hawaii Telescope. Down in Chile, at the VLTI, he is helping to develop integrated optics, which would combine beams efficiently on tiny silicon chips rather than in large, complicated rooms.
Although the technological hurdles to the OHANA project are still high, Perrin says that a bigger problem could be getting all the Mauna Kea observatories to simultaneously offer up their telescope time — a precious and fiercely guarded resource. "It will be easier to convince the communities that are behind the telescopes," says Perrin, "if we first demonstrate that interferometry is a big player in science today."
Kloppenborg, B. et al. Nature 464, 870-872 (2010).
Monnier, J. D. et al. Science 317, 342-345 (2007).
Perrin, G. et al. Science 311, 194 (2006).
Related external links
About this article
Cite this article
Hand, E. Telescope arrays give fine view of stars. Nature 464, 820–821 (2010). https://doi.org/10.1038/464820a