All imaging systems suffer from aberrations that reduce the resolution below the theoretical value often referred to as the diffraction limit. Although there are numerous sources of aberrations, the turbulent character of Earth's atmosphere is definitely the worst. In the last two and a half decades, a series of techniques to remove the detrimental effects of atmospheric turbulence have been developed and rigorously tested. These techniques are commonly referred to as adaptive optics (AO).1,2
The early success of AO for atmospheric compensation has lead researchers to apply it to other optical systems—imaging and non-imaging alike—such as medical imaging and high-power industrial lasers.3This surge in interest has been fueled by the use of new technologies like microelectromechanical systems (MEMS) and liquid-crystal devices. These have greatly reduced the complexity and cost of adaptive optical systems, allowing experimentation by more researchers.
A standard AO system typically has three main components: a wavefront sensor that measures the aberrations; a control system that reads the wavefront sensor and calculates and applies correction; and the wavefront corrective element1–3 The latter is the component that will physically change the impinging light in order to remove the aberrations. The most common type of corrective element is a deformable mirror, i.e. a flexible mirror with actuators (usually piezo-electric) that can deform the surface of the mirror to the phase conjugate of the aberrated wavefront, removing or greatly reducing the aberrations in the reflected light.
Such systems, while very successful in the past decades, have also proven extremely bulky, costly, and complex to operate. The use of MEMS as corrective elements has resulted in a reduction of cost, complexity, and weight. Standard AO systems require large optical benches, several racks of dedicated power supplies, and computer equipment. We developed a system with a physical form factor of only 51 × 71 × 51cm (20 × 28 × 20in), a total weight of 45kg (100lb), and with all power supplies and computer equipment in a single enclosure.4 This has allowed fully-functional AO systems to be added to several different telescopes with only a few days of setup time at each site.
On one such experiment, the AO system was transported roughly 300 miles and bolted to the back of a 1m telescope at the US Naval Observatory Flagstaff Station. The results shown in (Figure 1 were obtained after a setup lasting less than one day on the bright star α Lyrae (Vega). The left side shows the image of the star without the help of AO, while that on the right is with the aid of our portable system. The energy gathered by the telescope without AO is spread on a large area of the detector resulting in a dramatically-reduced signal-to-noise ratio (SNR). The image on the right shows more symmetry and a higher energy concentration, resulting in a higher SNR.
Figure 1. Image of the bright star α Lyrae (Vega) seen through a 1m telescope. The left side is the stellar image corrupted by the Earth's atmospheric turbulence. The right side is the same stellar image corrected by the AO system.
One way of measuring the amount of correction possible by an AO system is by measuring the so-called Strehl ratio. This is defined as the peak intensity of the aberrated stellar image over the peak intensity of the diffraction-limited image. Figure 2 shows a cross cut of the two stellar images overlapped, with the respective Strehl ratios. From this figure it is possible to determine that the AO system was able to produce a seven-fold increase in energy concentration, thus creating a higher SNR.
Figure 2. Cross cut and overlap of the two stellar images shown in Figure 1. The Strehl ratio is a measurement of the improvement attained by the AO system.
The integration of portable MEMS-based AO systems with standard astronomical telescopes allows atmospheric turbulence correction at a reduced cost and complexity. We have shown a seven-fold increase in Strehl ratio with a system that can be fully integrated on any telescope in two days or less. Future work will yield in a further reduction in size and power consumption by implementing new low-power electronics and computers, while increasing the Strehl ratio by integrating larger-aperture MEMS deformable mirrors.