Astronomy › Section F · Astronomical instruments › Topic F.2

Telescopes today

Today's giant telescopes have mirrors 8 to 10 metres across, sit on remote mountaintops and bend their own optics hundreds of times a second to cancel the blurring of the air. This topic explains how they are built, where they are placed and what limits how sharp their images can be.

From Palomar to the giants

The topic Telescopes showed why large research telescopes are reflectors. For decades the record holder was the 5-metre Hale Telescope on Palomar Mountain in California, completed in 1948. Its thick glass mirror weighed more than 13 tonnes and needed a massive steel frame to stop it sagging. Building much larger mirrors the same way seemed hopeless: a mirror twice as wide but just as stiff would be many times heavier. From about 1990, new technology broke the barrier, and more than a dozen telescopes with mirrors between 6.5 and about 10 metres across have been built since then.

Thin mirrors and active optics

The trick is to stop fighting the sag and correct it instead. Modern mirrors such as those of the 8.1-metre Gemini telescopes or the four 8.2-metre telescopes of the Very Large Telescope in Chile are only about 20 centimetres thick, far too thin to hold their shape alone. Computers measure the mirror's shape repeatedly while it observes and drive motorised supports, called actuators, that push on its back to keep the curve exact. This is called active optics. It lets a mirror much wider than Palomar's weigh not much more, held by a lighter, cheaper structure.

Segmented mirrors

Another approach is to build one large mirror out of many small ones. Each of the twin Keck telescopes on Maunakea in Hawaii, which began work in 1993 and 1996, has a 10-metre primary made of 36 hexagonal segments, each about 1.8 metres across. Actuators adjust every segment continuously so that together they act as one smooth surface. As of 2026, the largest single optical telescope working is the 10.4-metre Gran Telescopio Canarias on La Palma in the Canary Islands, also segmented. All the larger telescopes now under construction use segments, because a single mirror much beyond about 8.4 metres cannot be cast and transported.

Choosing a site

A large telescope costs hundreds of millions of dollars, so it must go where the sky is best. Earth's atmosphere causes four problems. Clouds and storms stop observing altogether. Water vapour absorbs infrared light, so dry sites are needed. Light pollution, city light scattered by the air, hides the faintest objects, and swarms of bright internet satellites in low orbit now streak across images too. And unsteady air blurs every image. The best sites are therefore high, dry, dark and steady: mountains in Chile's Atacama Desert, Maunakea at about 4200 metres, the Canary Islands and peaks in Arizona.

Seeing: why stars twinkle

The atmosphere is full of moving pockets of air, from centimetres to a metre or so across, each at a slightly different temperature. Each pocket bends light a little, like a weak lens, and the wind carries them through the beam many times a second. Starlight therefore arrives along a constantly shifting path, and the image dances and smears. To the eye this is the twinkling of stars. Astronomers call it seeing: bad seeing means strongly blurred images. Air that has flowed smoothly over a large ocean before meeting a mountain is especially steady, which is one reason island and coastal peaks make fine sites.

Measuring resolution

Resolution is the smallest detail an image can show, measured as an angle on the sky. Angles this small are given in arcseconds: one degree is 60 arcminutes and one arcminute is 60 arcseconds, so one arcsecond is \(\frac{1}{3600}\) of a degree. A 1-centimetre marble looks 1 arcsecond across from about 2 km away. Even with perfect air, the wave nature of light sets a limit, the diffraction limit:

\theta \approx 2.52 \times 10^5\ \frac{\lambda}{D}\ \text{arcseconds}

with wavelength \(\lambda\) and aperture \(D\) in the same units. Bigger apertures and shorter wavelengths give finer detail.

Worked example: Keck's sharpest view

Find the diffraction limit of a 10-metre Keck telescope in visible light at 500 nm, and compare it with good seeing of 0.5 arcsecond. 1. Write \(\lambda = 5.00 \times 10^{-7}\) m and \(D = 10\) m. 2. Substitute: \(\theta = 2.52 \times 10^5 \times \frac{5.00 \times 10^{-7}}{10} = 0.0126\) arcsecond. 3. Compare: \(\frac{0.5}{0.0126} \approx 40\), so even good seeing blurs the image about 40 times more than the telescope's optics alone would. 4. Repeat in the infrared at 2.2 micrometres: \(\theta = 2.52 \times 10^5 \times \frac{2.2 \times 10^{-6}}{10} = 0.055\) arcsecond. Check: the wavelength is 4.4 times longer, so the limit is 4.4 times coarser.

Adaptive optics

Ground telescopes can now partly beat the atmosphere with adaptive optics. A sensor measures how the incoming light has been distorted, using a bright star near the target or an artificial guide star made by shining a laser into the upper atmosphere. A computer then reshapes a small, flexible mirror in the light path, hundreds to more than a thousand times a second, to undo the distortion. Adaptive optics works best in the infrared, where the needed corrections are slower and smaller. With it, the largest ground telescopes reach about 0.1 arcsecond or better, as sharp as the Hubble Space Telescope in visible light.

Worked example: steady air or not?

An astronomer wants to separate two stars 0.2 arcsecond apart with an 8.2-metre telescope at a wavelength of 1.6 micrometres. 1. Diffraction limit: \(\theta = 2.52 \times 10^5 \times \frac{1.6 \times 10^{-6}}{8.2} = 0.049\) arcsecond. 2. The optics alone could easily split the pair, since 0.049 is well under 0.2. 3. Seeing at an excellent site is about 0.5 arcsecond, larger than the separation, so without correction the two stars blur into one. 4. Conclusion: the pair is resolved only with adaptive optics. Check: the limit is about four times smaller than the separation, so a well-corrected image should show two clear points.

Common mistakes to avoid

First, thinking a bigger telescope on the ground always gives sharper images; without adaptive optics, seeing limits it to a few tenths of an arcsecond. Second, confusing active optics, which slowly holds a large mirror's shape, with adaptive optics, which rapidly cancels atmospheric blurring. Third, mixing units in the diffraction formula; \(\lambda\) and \(D\) must be in the same unit, such as metres. Fourth, assuming longer wavelengths give finer detail; they give coarser detail. Fifth, picking sites only for darkness; dryness and steady air matter as much.

Key takeaways

Since about 1990, thin actively supported mirrors and segmented mirrors have made 8- to 10-metre telescopes possible; Keck uses 36 hexagonal segments, and the 10.4-metre Gran Telescopio Canarias is the largest working single optical telescope as of 2026. The best sites are high, dry, dark and steady. Turbulent air causes twinkling and blurred seeing. Resolution is an angle, often in arcseconds, with a diffraction limit θ ≈ 2.52 × 10⁵ λ/D. Adaptive optics reshapes a flexible mirror rapidly to reach about 0.1 arcsecond from the ground.

Flashcards

Flashcard

Hale Telescope

The 5-metre Palomar reflector of 1948

Heavy mirror, massive frame.

Test your understanding

Question 1 of 5

What is the diffraction limit of an 8.0-metre telescope at a wavelength of 1.0 micrometre? (θ ≈ 2.52 × 10⁵ λ/D arcseconds)

Question 2 of 5

Which site would be best for a new infrared telescope?

Question 3 of 5

What does adaptive optics correct?

Question 4 of 5

How does each Keck telescope achieve its 10-metre aperture?

Question 5 of 5

Why do stars twinkle when seen from the ground?

← Previous topicNext topic →