Telescopes, Computer-Controlled Mirrors

The first half of the twentieth century saw a phenomenal increase in the size of astronomical telescopes, culminating with the 5-meter Hale instrument on Mount Palomar. However by midcentury it appeared that the ultimate size of earth-bound telescopes was inherently limited. Despite the use of new materials and advanced engineering, increasing the size of telescope mirrors seemed to make them inherently more sensitive to flexure and vibration.

Moreover, the resolving power of earth- bound telescopes seemed to be ultimately limited by the distorting effects of atmospheric turbulence. One of the primary justifications for funding the Hubble space telescope was to avoid these difficulties.

In the second half of the twentieth century astronomers explored new technological approaches to solving these problems, many of which concentrated on developing increasingly sophisticated sensing devices. But there was also an impetus to rethink the telescope from its traditionally passive role and to give it some capacity to react to changes in viewing conditions. Eventually two separate, but complementary, systems emerged: active optics and adaptive optics.

Active optics involves changing the shape of the telescope’s primary (and in some cases, secondary) mirror to optimize the instruments’ resolving power (see Figure 1). Active optic systems analyze the image of a reference object (typically a bright star located near the object being studied) to generate a series of changes in the telescope’s optical system, optimizing the image of the object being viewed.

Figure 1. Active optics system

These changes may be performed relatively slowly and can correct for a variety of problems, including low-frequency wind vibrations, uneven thermal expansion, gravitational distortion and even the instrument’s permanent figuring errors. Instruments with active optical systems can be made significantly lighter than conventional telescopes and are well suited to use on mirrors made up of multiple segments.

Adaptive optics involves changing the shape of a much smaller mirror, typically positioned within the telescope’s optical train, with the goal of correcting wave-front distortions that occur as light passes through the earth’s atmosphere (see Figure 2). Adaptive optics uses a guide-star feedback loop similar to that used in active optics but with much smaller corrections that are ideally made in the space of a few milliseconds.

Figure 2. Adaptive optics system

Although suggested as early as 1953, practical implementation of these systems required the development of a variety of supporting devices, including wave-front sensors, deformable mirrors, actuators, computers, and computer software. While active optics is now considered a mature technology, the routine employment of adaptive optics to the entire visual spectrum is still under development. Significant correction of the atmospheric distortion of infrared light has been successfully achieved, but correction of distortion to higher frequencies remains a challenge.

Another important challenge has involved finding suitable reference stars for image and wavefront correction. The ideal reference is a bright star located fairly close to the object being studied but far enough away from the telescope’s optical axis that its light is not part of the observation. The source needs to be bright enough to overcome noise in the wave-front detectors and it needs to be close enough to the science object to have roughly the same light-path through the atmosphere. Finding a suitable natural guide star (NGS) for every observation has proved to be a problem and the possibility of generating artificial reference stars was an early area of investigation.

This work received an unexpected boost from the U.S.’s Strategic Defense Initiative (SDI) in the 1980s. SDI planned to use a form of adaptive optics to focus a destructive laser beam on incoming missiles. To permit rapid response, the system used an artificial reference star created by shining a laser on a naturally occurring layer of sodium atoms approximately 90 kilometers above the earth’s surface. When the SDI program was cancelled in 1991 much of this technology was turned over to astronomers. The use of laser guide stars (LGS) in astronomy dates from this time.

While the use of LGS appears to solve the reference star problem, the relative nearness of these objects introduces the problem of conical anisoplanatism. This refers to the fact that light from the LGS arrives at the telescope in a path that is cone shaped. This can cause the LGS wave-front distortion to differ significantly from that of astronomical object being studied. LGS systems can also be expensive to install and operate. A number of alternatives have been proposed, including the use of multiple guide stars (either natural or artificial) to produce a tomographic AO correction for entire sections of the sky.

Progress in both active and adaptive optical systems continues to accelerate and it is clear that future earth-bound astronomical telescopes will depend heavily on them. While complete correction of observing errors is not possible, even partial correction can produce dramatically improved images. The image quality of modern astronomical telescopes is commonly expressed in terms of the strehl ratio, which is defined as the ratio of the peak intensity of an image divided by the peak intensity of a diffraction-limited (ideal) image. A perfect strehl ratio is 1.0; for reference, the Hubble space telescope (HST) is estimated to have a strehl ratio of 0.97.

For earth-bound telescopes, the highest achievable strehl may be around 0.8 and day-to-day observations may fall short of that goal. However it has already been demonstrated that, using adaptive optics, a 10-meter surface telescope can produce infrared images that surpass those produced by Hubble.

This is because of the much greater light-gathering area of the surface telescope. Moreover, the cost of building and operating large earth-based instruments can be significantly lower than launching telescopes into space. With the size of launchable space telescopes currently limited, attention is being increasingly directed towards the construction of large AO earth based observatories. The next generation of visual telescopes may be in the 20 to 30 meter range and instruments as large as 100 meters have been proposed.