Telescope Technology Clears the Air For Astronomers
Newly developed active and adaptive optics give sharper views, compensating for mirrors' shortcomings, atmospheric distortions
Kitt Peak, Ariz.
WITH lasers to light the way and mechanical pistons to gently massage the sharpest images of stars and galaxies out of a new generation of telescope mirrors, astronomers are calling on two cutting-edge technologies to revolutionize ground-based optical and infrared astronomy.
The goal: to bring Hubble Space Telescope quality to observatories on mountaintops around the world.
Within a 40-mile radius, two facilities illustrate the trend: the 3.5-meter WIYN telescope here on Kitt Peak, and the Whipple Observatory's Multiple-Mirror Telescope (MMT) on Mt. Hopkins, also in Arizona.
Taken together, these observatories are using techniques known as active and adaptive optics. Active optics ensure that new lighter-weight telescope mirrors maintain their painstakingly engineered shape as the telescope tracks objects in the night sky. Adaptive optics are designed to remove the distortion in starlight that comes as the light passes though turbulence in Earth's atmosphere.
Individually, these approaches - both still in their development stages - have brought significant improvements to telescopes. Together, they will allow mountaintop telescopes to begin exploring in unprecedented detail objects ranging from distant galaxies to nearby regions where stars are forming - and at a substantially reduced cost.
``A telescope works only as well as its weakest link,'' says Blaire Savage, an astronomy professor at the University of Wisconsin at Madison. ``The basic drive for active optics has been one of pushing the technology for supporting larger mirrors. Combined with this is the fairly recent realization that you can beat the effects of the atmosphere.'' This, Dr. Savage says, makes efforts to improve mirrors, their supporting structures, and the telescopes' operating environment all the more important.
Although astronomers first posted the concept for removing the twinkle from starlight in the 1950s, the concepts were deemed beyond the technological reach of the day. Astronomers had to look for other aspects to improve.
Concepts for active optics began to evolve in the late 1970s, Savage says, ``when it became clear that you could do a lot more in the observatory itself'' to boost image quality with innovations in mirrors and the structures that support them. These innovations were being driven by the need for larger telescopes to observe ever-fainter objects while holding down costs.
One mirrormaking technique, developed by Roger Angel at the University of Arizona's Optical Sciences Center, involves melting glass blocks over a honeycomb form in a large rotating oven. The spin distributes the molten glass and gives the mirror its concave shape; the honeycomb style leaves it lighter in weight and enables it to adjust more rapidly to changes in air temperature, another factor that affects image quality.
The downside: Lighter mirrors tend to sag more than their solid-glass counterparts as their position changes. Although the sag is measured in thousandths of an inch, it is enough to distort the image. Hence the need for actuators to constantly tug and nudge the mirror as it tracks objects across the night sky. The idea is to keep the mirror's shape as close as possible to the design ideal, Savage says, nodding toward the 66 actuators protruding from the back of the WIYN telescope.
The instrument is the result of a collaboration between the University of Wisconsin, Indiana University, Yale University, and the Tucson-based National Optical Astronomy Observatories. (Hence the WIYN acronym.)
Active optics, combined with elaborate temperature controls to reduce distortions near or within the main mirror itself, gives WIYN an ability to distinguish between closely spaced objects that significantly exceeds that of Kitt Peak's older 4-meter Mayall telescope.
``Active optics are required on every new-style telescope,'' notes Laird Thompson, an astronomy professor at the University of Illinois at Champaign-Urbana. Among them: the 8-meter Gemini telescopes now under construction in Hawaii and Chile, as well as the MMT's upgrade to a 6.5-meter, single mirror telescope.
Seeing with `no atmosphere'
If active optics can vastly improve a telescope's resolution, another measurement suggests why astronomers are pushing to use adaptive optics as well. A telescope's ability to distinguish closely spaced objects is measured in terms of its angular resolution, which uses degrees, minutes, and seconds of arc to establish spatial relationships between distant astronomical objects. WIYN's optical system boasts an average angular resolution of 0.7 arc-seconds on Kitt Peak. In the lab, its ``no atmosphere'' resolution increases to 0.04 arc-seconds, a more than 10-fold improvement and comparable to the Hubble Space Telescope's performance in orbit.
Adaptive optics - nurtured by the Pentagon for satellite surveillance and the Strategic Defense Initiative and declassified for wider use in 1991 - holds the promise of bringing telescopes closer to that no-atmosphere ideal.
In space, stars are so distant that their light waves arrive at Earth with an essentially straight wavefront, much like that of a line of surf at a beach. But as the light pierces the atmosphere, it encounters turbulence that jumbles what once was a tidy image. When they reach the telescope, tightly spaced points of light are rendered as fuzzy patches.
An adaptive-optics system needs a reference light source high enough to pass through all the distortion layers of the atmosphere. If a star of sufficient magnitude is near enough to the object of interest, light from the ``guide star'' can be used. But turbulence limits the usefulness of natural guide stars, researchers say, particularly at visible wavelengths. A star must be at least 10th magnitude in order to serve as a reference for adaptive optics. Dimmer stars do not give enough light to provide accurate sampling at the speeds and across the individual zones of the mirror to be corrected.
Yet 10th-magnitude stars are relatively few compared with the number of objects astronomers want to study. This has led several groups to experiment with so-called laser guide stars. Sodium lasers, for example, have been used to generate a glowing spot in a layer of atomic sodium that lies about 100 kilometers (about 60 miles) above earth. Mounted either on the telescope or close by, such lasers can provide a reference anywhere in the sky.
Dr. Thompson and colleagues, for example, are using a sodium laser with the 2.5-meter telescope at Mt. Wilson in California to experiment with adaptive optics.
Adaptive optics use computer-driven, high-speed actuators to bend a thin deformable mirror to correct the distortion as the mirror reflects light to the telescope's viewing instruments.
The movements are tiny, measured in microns; yet depending on whether the system is designed for longer-wavelength infrared observation or shorter wave-length visible-light work, the actuators can number in the hundreds, and the adjustments must be made hundreds of times a second. The corrected light is constantly sampled; new information about the wavefront's shape is constantly fed to the computer to readjust the deformable mirror.
Despite the enthusiasm for adaptive optics, obstacles remain, researchers say.
One is laser reliability. ``Lasers have a reputation for being finicky,'' says Claire Max, a physicist at Lawrence-Livermore National Laboratory in California who has been involved in classified and unclassified adaptive-optics research. ``To have a laser fail while an astronomer is in the middle of making an observation would be grounds for justifiable homicide,'' she jokes.
Laser power is another issue. The University of Arizona's Dr. Angel has been working with a team to design an adaptive-optics system for the MMT, which is being upgraded to a single 6.5-meter telescope. He says that while a 5-watt laser is sufficient for infrared work, designing adaptive-optics systems for visible light requires lasers with output powers of up to 100 watts.
``That's at the cutting edge of laser technology,'' he notes, adding that on 8- to 10-meter class telescopes, several lasers are required to assure correction out to the edges of the main mirror.
Moreover, the demand for computing power grows as a system moves from infrared to visible light and as systems are designed for ever-larger telescopes. ``You can't use off-the-shelf systems anymore,'' Dr. Max says.
Yet the payoffs are huge. Says Angel, ``A 6.5 meter telescope fully corrected at a wavelength of 2 microns [in the infrared] essentially matches [that of] the Hubble Space Telescope at visible wavelengths.''
And that's without ever leaving the ground.