Adaptive optics using multiple laser guide stars

The primary limitation to taking sharp, high-resolution images of the heavens with optical and near-IR telescopes is the scattering of the incident light by the Earth's turbulent atmosphere. Similar to air rising from the ground on a hot day causing a heat mirage, hot and cooler air mixing above a telescope blurs the resulting images. Astronomers refer to this effect as ‘seeing,’ and—depending on weather conditions—it can reduce image resolution from 0.5 to 1 arcsecond at visible wavelengths. In the absence of the Earth's atmosphere, a telescope's resolution is set by its size, e.g., at 0.02 arcseconds for a 6.5m visible-light telescope.

Adaptive optics (AO) systems can compensate for much of the degrading effects of the atmosphere. Until very recently, these systems relied on either a single bright star in the sky, or on an artificial star generated by a powerful laser, to measure atmospheric fluctuations. This system limits the best optical correction to a single point on the sky and correction quality degrades away from that location. However, this deficiency can be overcome by using multiple stellar or laser sources, thus enabling many types of wide-field AO correction.¹ The simplest of these correction methods, ground-layer adaptive optics (GLAO), has been suggested as a way to improve wide-field imaging for large telescopes.² Atmospheric-turbulence measurements from multiple guide sources located far from each other (2–10+ arcminutes) can be averaged to estimate turbulence close to the telescope aperture.

Half to two-thirds of the total atmospheric turbulence is found within the first kilometer above a telescope, which is common to every point on the sky. Correcting for only this ground-layer turbulence will therefore lead to a dramatic improvement in imaging over the full field of view facilitated by the guide sources. In addition, by employing laser guide stars that are always pointed in the same direction as the telescope, it is no problem if there are not enough bright natural guide stars available to properly measure atmospheric turbulence.

Figure 1. The multiple laser-guide-star adaptive-optics (AO) system in operation at the 6.5m MMT (formerly the multiple-mirror telescope) in southern Arizona. (Images courtesy Thomas Stalcup.)

Our team has deployed the world's first multiple laser AO system at the 6.5m MMT (formerly the multiple-mirror telescope) in southern Arizona (see Figure 1) to demonstrate GLAO and other advanced AO techniques.³ The MMT laser-AO system comprises a laser-launch telescope that projects five Rayleigh laser guide stars on a 2 arcminute diameter pentagon, an adaptive secondary mirror, a Cassegrain-mounted wavefront sensor, and a PC-based real-time controller. The system measures ground-layer turbulence and produces corrections to the adaptive secondary mirror 400 times per second, which was demonstrated for the first time during initial commissioning in February 2008.⁴

Figure 2 compares a stellar image taken with and without the system. At a near-IR wavelength of 2.14μm, the uncorrected image has a full width at half maximum (FWHM) of 0.70 arcseconds. However, with GLAO correction, the FWHM decreases to 0.33 arcseconds and the peak intensity increases by a factor of 2.3. Furthermore, in terms of better 'seeing,' a somewhat worse-than-median night was improved by the GLAO system to excellent conditions, normally only experienced once every 25 nights.

Figure 2. Comparison of stellar images in the near-IR without (left) and with (right) ground-layer AO (GLAO) correction. With GLAO, the image width is reduced from 0.70 to 0.33 arcseconds and the peak intensity is increased by a factor of 2.3. Each grid point represents 0.107 arcseconds on the sky.

Figure 3 shows the wide-field imaging improvement afforded by GLAO in the near-IR. Over a 110 arcsecond field, an uncorrected star's average FWHM improves from 0.72 to 0.58 arcseconds with the GLAO system. Unfortunately, technical problems in the control loop hindered the system's ultimate performance, but have since been corrected and future GLAO observations should not be similarly affected.

Figure 3. A science-camera field showing four times magnified images of stars without (left boxes) and with (right boxes) GLAO correction over a 110×110 arcsecond field of view. Average stellar image widths are decreased by 19% with GLAO correction. The magnified images have been rescaled for clarity.

We have shown that the MMT's laser-GLAO system reduces the atmosphere's blurring effects over wide fields of view, demonstrating the reduction of stellar image widths by as much as 53% in the near-IR, and uniform correction over a 110 arcsecond field of view. We expect the system to deliver 0.1–0.2 arcsecond images in the near-IR during median seeing over the coming months as we finalize commissioning. The lessons learned from this system will be invaluable for our future research to build the multiple laser-AO systems required for the next generation of extremely large telescopes, such as the 25.4m Giant Magellan Telescope and the Thirty Meter Telescope.