Astronomers trying to get sharp images are face with a dilemma. They can make their mirrors larger, which improves the angular resolution at the limit of the optics linearly with the size of the mirror. The problem is that for any telescope bigger than about 30 centimeters, the limit on image sharpness and angular resolution is the blurring of incoming starlight by the Earth's atmosphere, not the telescope optics. So the telescope aperture is in part wasted. Or they can go to the enormous expense of launching a telescope into space, which costs 10-20 times more that building the same sized telescope on the ground.
Adaptive optics is a way forward that avoids this dilemma. It works like this. Light from an astronomical object travels in straight lines through the vacuum of space so the wave front, or set of waves that left the object at the same time, defines a flat surface. But as the light travels through the atmosphere, its speed depends slightly on density, and the turbulent motions of air in the upper atmosphere create slight variations in density. So the smooth incoming wave front gets distorted into a corrugated sheet. Like a funhouse mirror, that creates distorted images. The turbulent motions are rapid so this distortion of the light changes many times a second. The net result is a blurred image. Adapive optics has two key ingredients. One is a wave front sensor, to detect exactly how the incoming wave front has been distorted by the atmosphere. That is done with by using a point source near the object to be studied, as a reference, or by bouncing a laser off the same atmospheric layer that causes the distortion, and analyzing the collected laser light. The second ingredient is a flexible mirror that can have its shape rapidly adjusted to exactly compensate for the distorted wave front. Since the primary mirror is usually too large and rigid to bend, the deformable mirror is the smaller secondary mirror. If this sounds difficult, it is! Adaptive optics spend more than a decade in the lab and another decade being tested on telescopes before it worked.
Today, every major telescope either has some version of adaptive optics operating, or it is under development. It was first perfected at the longer wavelengths of the near infrared, where the technical challenges are less. Now it works at optical wavelengths too. The proof of the pudding is, as they say, in the eating. The best images made with adaptive optics (abbreviated AO) recover the diffraction limit of the telescope optics. In 2013, the adaptive optics system developed at the University of Arizona for the Magellan 6.5-meter telescope in Chile made the sharpest image of the night sky ever. It had an angular resolution of 0.02 arc seconds, good enough to see a dime a hundred miles away. With the best images from the ground exceeding the image quality of the Hubble Space telescope, some of the advantage of space astronomy has disappeared. Once again, technical innovation is pacing the rate of discovery in modern astronomy.