Adaptive Optics: Taking the Twinkle Out of Stars

by Louise Good

Adaptive Optics (AO) System (left): Light from the telescope is sent to a deformable mirror, then to a beamsplitter, where part of the light is reflected to the wavefront sensor. The wavefront sensor measures the distortion in the wavefront and sends a correction signal to the deformable mirror. The deformable mirror changes shape to remove the distortions in the lightwave before the light goes to the camera. Adapted from M. McInnis, LLNL, and C. Max, UCO/Lick Observatory.

Adaptive optics on and off (right): Hokupa`a-85 took these images on the Gemini South telescope in January 2005. With the AO system off, the field looks like one large star, but with the AO system on, it is possible to see that there are two stars, a bright one and a faint one (a binary system). To achieve this resolution, the AO system must measure the wavefront, calculate correction voltages, and apply these voltages to the deformable mirror about 1,000 times per second.

Temperature fluctuations in Earth's atmosphere act like small, randomly sized and oriented weak lenses that cause stellar images to degrade and dance (twinkle), limiting the resolution and sensitivity of ground-based telescopes. "Seeing," as these effects are called, varies with the site and conditions but never vanishes. The only way to avoid it is to launch a telescope into space. Mauna Kea and, to a lesser extent, Haleakala have better seeing than most observatory locations, yet even at these exceptional sites, the atmosphere turns pinpoint sources of light (such as stars) into slightly fuzzy blobs.

The Hubble Space Telescope (HST) was conceived and launched in an attempt to get around this problem by putting a telescope above the atmosphere. But even small telescopes in space—HST's mirror is only 2.4 meters (7.9 feet) in diameter—are very expensive. An 8-meter (26-foot) telescope on Earth has more than 10 times the collecting area of HST, so it can collect 10 times the amount of light from distant objects. The challenge is to free this light from atmospheric distortions.

Adaptive optics (AO) systems were first developed by the defense industry in the late 1970s and early 1980s to image artificial satellites. In the late 1980s, astronomers started experimenting with AO technology to restore the performance inherent in large ground-based telescopes. Just as noise-canceling headphones use microphones to sample the surrounding sound and then cancel it out by creating sounds that are the sonic opposite, AO systems use a wavefront sensor to determine what the atmospheric distortions are and apply an equal but opposite distortion to correct the wavefront before the image reaches the camera attached to the telescope. The key components of any AO system are a wavefront sensor to measure the incoming light and a deformable mirror to correct it. Most AO systems in the world measure the slope of the incoming wavefront and then use push-pull actuators behind a thin mirror to correct it.

The IfA has been a leader in adaptive optics research and development for more than 15 years. In 1988, François Roddier, who originally led the IfA AO Group, discovered a new way to measure starlight distortion and showed that when coupled to a different type of correcting element, a very powerful, very simple AO system could be built. These curvature adaptive systems were developed at the IfA and have slowly grown in use around the world.

Roddier retired in December 2000. The current AO group led by Christ Ftaclas and Mark Chun arrived in January 2002. Group members now include engineer Peter Onaka and assistants Sharon Velez Erickson and Sarah Cook.

Left: A curvature deformable mirror like that used in the adaptive optics system Hokupa`a-85. Right: The 85-element lenslet array that is the heart of the wavefront sensor. When a beam of light falls on the array, each individual element focuses the light on a spot at the back of the array, where it enters an optical fiber. Each of the 85 fibers is connected to a photon-counting detector. Hokupa`a, which means "immovable star," is the Hawaiian name for the North Star.

Their main focus has been designing and building Hokupa`a-85, an 85-element successor to the original Hokupa`a (used to commission the Gemini North telescope in June 1999), and the AO system for the Gemini Near Infrared Coronagraphic Imager (NICI) for Gemini South. Hokupa`a-85 was commissioned on the Gemini South telescope in Chile, but will return to IfA Manoa for further refinements. According to Ftaclas, the AO Group is now "trying to move from proof of concept and modest improvements in image quality to truly high-performance levels." This will involve building systems with hundreds of elements. Such a goal "requires components, especially curvature deformable mirrors, that are simply not available now. We are one of the few groups in the world researching such components, but we have yet to demonstrate deformable mirrors that work the way the model says they should. This is a quality control issue, and we are learning at each step what can go wrong, how we can prevent it, and how we can make it better. Hokupa`a-85 taught us a lot, and we are optimistic about the AO system for NICI," which will be used to search for extrasolar planets. He adds, "After four years of designing and building instruments and systems to meet deadlines, we are just now managing to run our first complete laboratory characterization of one of our curvature deformable mirrors."

In addition to building high-end, very expensive systems, Ftaclas hopes to be able to build smaller, low-cost (less than $100,000) systems for smaller campus-type telescopes through four-meter systems. He says, "Just as better detectors improve all telescopes, low-cost AO systems improve the entire astronomical infrastructure."