Adaptive Optics: How Telescopes Cancel Out Earth’s Atmosphere

Target Pillar: Observing the Cosmos: Telescopes, Instruments, and Data

Last Updated: September 28, 2026 Author: Sarah Mitchell

Introduction

Stars twinkle because the air above us is never still. Pockets of warmer and cooler air constantly bend starlight on its way down, which is charming when you’re stargazing and a serious problem when you’re trying to photograph a distant planet. For most of the history of astronomy, that blur set a hard limit on what ground-based telescopes could see, no matter how large their mirrors were. Adaptive optics is the technology that broke through that limit. Instead of avoiding the atmosphere, as the Hubble Space Telescope does by orbiting above it, adaptive optics measures the distortion as it happens and cancels it out, many times every second.

The Basic Idea: Measure the Blur, Then Undo It

Every adaptive optics system, whatever telescope it’s attached to, has to do the same three jobs. It needs a reference light source whose true appearance is known, a sensor that measures how the atmosphere has distorted that light, and a mirror that can change shape fast enough to compensate.[^1]

The reference is usually a bright star close to whatever the astronomers actually want to study. Because the star should look like a perfect point of light, any smearing in its image tells the system exactly how the local atmosphere is bending light at that moment.[^2] A wavefront sensor reads that distortion, a computer works out the correction, and a deformable mirror, a thin, flexible mirror pushed and pulled from behind by many small actuators, reshapes itself to cancel the error before the light reaches the science camera.[^3]

The speed involved is what makes this remarkable. Atmospheric turbulence changes on timescales of milliseconds, so the whole measure-and-correct loop has to run hundreds or thousands of times per second. The adaptive mirror planned for Europe’s Extremely Large Telescope, for example, will adjust its shape about 1,000 times per second, to an accuracy of a few tens of nanometers.[^1]

What It Looks Like at a Real Observatory

The twin 10-meter Keck telescopes in Hawaii are among the best-documented examples. The Keck I system uses a Shack-Hartmann wavefront sensor, which splits the incoming light across a grid of 20 by 20 tiny lenses to map the distortion, feeding a deformable mirror with 349 actuators.[^4] An earlier peer-reviewed description of the Keck laser system quoted sensor readout rates of up to 670 times per second, though the hardware has been upgraded since, so treat that figure as a snapshot of the original design rather than today’s specification.[^5]

Under good conditions, Keck’s system delivers near-diffraction-limited images in the near-infrared, meaning the pictures approach the theoretical sharpness limit for a 10-meter mirror. Researchers describing the system report Strehl ratios (a standard measure of how close an image comes to that perfect limit) on the order of 30 to 40 percent in the infrared K-band.[^6] That may sound modest, but it represents a dramatic improvement over uncorrected ground-based images.

The Reference Star Problem, and the Laser Solution

There’s a catch in the basic recipe: the reference star must be bright, and it must be very close to the target on the sky. Suitable natural stars simply aren’t available for most of the night sky.[^2]

Astronomers solved this by making their own star. A laser tuned to the sodium wavelength is fired up through the telescope, where it excites sodium atoms in a layer of the upper atmosphere and creates a glowing spot that acts as an artificial “laser guide star.”[^7] Because the astronomers control where the laser points, they can place this artificial star right next to almost any target. ESO notes that thanks to laser guide stars, almost the entire sky can now be observed with adaptive optics.[^2]

Laser guide stars aren’t perfect substitutes. A laser spot can’t tell the system about the overall wobble of the image, so many systems still need a faint natural star for that one correction, and a spot created at a finite altitude doesn’t sample the atmosphere in exactly the same way starlight does.[^5] Newer designs address this with multiple lasers. ESO’s Extremely Large Telescope will use up to eight, and Keck’s upgraded system uses laser tomography, combining several laser beacons to reconstruct the turbulence in three dimensions, a technique researchers describe as crucial to the success of the next generation of extremely large telescopes.[^1][^4]

Where Adaptive Optics Still Struggles

Adaptive optics is powerful, but it has real limits. It works best at infrared wavelengths, where turbulence is less severe, and performance falls off quickly at shorter, visible wavelengths.[^6] The corrected region of sky is also small: because the correction is tuned to the light path toward one reference, only a narrow field around it comes out sharp.[^3] Wider-field designs exist, including a proposed Keck upgrade aimed at correcting a one-arcminute field at visible wavelengths, but those systems require several deformable mirrors, multiple lasers, and much more computing power.[^6]

The engineering demands are steep. ESO lists the requirements bluntly: very fast, low-noise detectors, powerful and reliable sodium lasers, processors handling billions of operations per second, and deformable mirrors with thousands of actuators responding at kilohertz speeds.[^3] That’s why adaptive optics arrived on major professional telescopes gradually rather than all at once.[^8]

Ground Versus Space: Complements, Not Rivals

It’s tempting to frame adaptive optics as ground telescopes catching up to space telescopes, but the two approaches have different strengths. A space telescope sees ultraviolet light the atmosphere blocks and delivers a clean, stable image across a wide field. A large ground telescope with adaptive optics can carry a mirror far bigger than any launched to date, collecting far more light and resolving finer detail within its small corrected field. Astronomers routinely combine both, using Hubble and the James Webb Space Telescope alongside adaptive-optics-equipped ground observatories on the same targets.

Frequently Asked Questions

Sources

1. ESO — Adaptive Optics (Extremely Large Telescope)
2. ESO — Adaptive Optics
3. ESO — What Is Active and Adaptive Optics?
4. arXiv — On-Sky Capabilities and Performance of the Keck All Sky Precision Adaptive Optics System
5. Publications of the Astronomical Society of the Pacific — The W. M. Keck Observatory Laser Guide Star Adaptive Optics System: Overview
6. arXiv — Performance Simulations for KOLA: High-Resolution, Visible-Light AO Correction Over a 1 Arcminute Field
7. Publications of the Astronomical Society of the Pacific — The W. M. Keck Observatory Laser Guide Star Adaptive Optics System: Performance Characterization
8. Astronomy.com — How Does Adaptive Optics Work?

Note on methodology: technical details above are drawn from ESO’s published documentation and peer-reviewed and preprint papers on the Keck adaptive optics systems. Some Keck specifications come from a 2006 overview and have since been upgraded; system performance figures vary with observing conditions. Consult the observatories’ own pages for current specifications.