Telescopes
How We See Far
A Telescope Is a Bucket
A telescope’s job is not really to magnify. Magnification is trivially easy and almost never the limiting factor. The real job is collecting light, and a telescope is best understood as a bucket held out in a very light drizzle of photons.
A faint galaxy might deliver a handful of photons per second to a patch of ground the size of your hand. Your eye’s pupil, at its widest, is about seven millimeters across. A modern research mirror is eight meters. That is over a million times the area, which is over a million times the photons, which is the difference between an empty sky and a spiral arm.
This is why every advance in astronomy has been an advance in aperture, and why telescopes are named by their diameter rather than their power. It is also why exposure time matters so much. A long exposure is a small bucket left out for longer, and for a fixed amount of observing time, area is the resource you cannot substitute for.
Why Every Big One Is a Mirror
The first telescopes bent light through a lens. Every large telescope built in the last century bounces it off a mirror instead, and there are three separate reasons, any one of which would have been enough.
First, a lens bends different colors by different amounts, so a simple lens brings blue to a focus in a different place from red and every star wears a colored halo. Mirrors reflect all colors identically and the problem simply does not arise. Second, a lens must be supported at its rim and light has to pass through it, so it sags under its own weight and any flaw in the glass matters. A mirror is supported across its whole back and only its front surface has to be right. Third, glass absorbs, and a thick lens for a large telescope would swallow much of what it gathered.
Even mirrors run out of room. Above roughly eight meters a single disc of glass cannot be cast, transported, or supported without deforming. The answer is segmentation: build the mirror from dozens of hexagons, each on its own actuators, and adjust them continuously so they behave as one surface. The Webb telescope’s eighteen segments were folded to fit inside a rocket and unfolded in space, which is why its mirror is a honeycomb rather than a disc.
The Limit Diffraction Sets
Even a perfect mirror in perfect conditions cannot produce a perfect point. Light is a wave, and a wave squeezed through an opening spreads. A star, however far away and however small, always arrives as a small disc surrounded by faint rings.
Two stars close together produce two overlapping discs, and if the discs overlap enough you see one blur. Widening the aperture shrinks the discs, and at some width the two separate. That threshold is the diffraction limit, and it depends on exactly two things: the diameter of the opening and the wavelength of the light. Nothing about the quality of the glass enters into it.
The wavelength dependence has a real consequence. The same telescope resolves finer detail in blue light than in red, and considerably less in the infrared. That is one of the costs the Webb telescope pays for seeing the early universe – it works at long wavelengths, so despite a mirror far larger than Hubble’s, its sharpness in the infrared is comparable rather than overwhelming.
What the Air Does to the Image
For most of astronomy’s history the diffraction limit was irrelevant, because the air got there first. Pockets of slightly different temperature drift across the line of sight, each bending light a little differently, and the image dances and smears many times a second. Astronomers call it seeing, and at a good site it blurs everything to about an arcsecond – roughly what a twelve centimeter telescope would manage on its own. A ten meter mirror gained no sharpness at all over a small one. It only gained light.
Adaptive optics broke that ceiling. A sensor measures the distortion hundreds or thousands of times a second, and a small deformable mirror bends in the opposite shape to cancel it. The correction has to be recomputed faster than the air changes, which is why this had to wait for fast computing.
The system needs a bright reference point near the target to measure against, and most of the sky has none. So observatories make their own. A laser tuned to a wavelength that sodium atoms absorb is fired upward, and it excites a thin layer of sodium about ninety kilometers up, left over from vaporized meteors. That patch glows, and it becomes an artificial star to correct against. Ground telescopes using this now beat Hubble’s sharpness at some wavelengths.
Faking an Enormous Aperture
Since sharpness depends on the width of the opening, there is a trick available. Two telescopes some distance apart, with their light combined so the waves interfere correctly, resolve detail as finely as a single mirror the size of their separation. They collect only as much light as their actual mirrors, so they are dim, but they are extraordinarily sharp.
The combination has to preserve the light’s phase, which means the path lengths must be matched to a fraction of a wavelength while the Earth turns. At visible wavelengths this is brutally hard and works over hundreds of meters. At radio wavelengths, which are millions of times longer, the tolerance is far more forgiving, and radio interferometry has spanned continents for decades.
The Event Horizon Telescope took this to the limit by recording data at observatories on several continents with atomic clocks, shipping the drives to one place, and combining them afterwards. The effective aperture was the diameter of Earth. That is what produced the images of the shadows around the black holes in Messier 87 and at the center of our own galaxy.
Leaving the Air Behind
Adaptive optics fixes blurring but not blocking. The atmosphere is opaque across most of the spectrum: ultraviolet, most infrared, X-rays, and gamma rays never reach the ground at all. Whole branches of astronomy are only possible above it, which is the real argument for space telescopes rather than image quality.
Webb’s design follows from one decision. To see the first galaxies you must look in the infrared, because the expansion of space has stretched their light far out of the visible range. But everything warm glows in the infrared, including the telescope itself, so the instrument has to be colder than what it is looking for. Hence the shield the size of a tennis court, hence the orbit one and a half million kilometers out where the Sun, Earth, and the Moon all sit in the same direction and one shield blocks all three, and hence the operating temperature around forty degrees above absolute zero.
The Limits That Remain
Diffraction is not negotiable. There is no processing trick that recovers detail finer than the aperture allows, because the information was never collected. Software can sharpen an image within that limit and cannot invent anything beyond it, which is worth remembering whenever an image is described as enhanced.
There is a second floor beneath that one. Light arrives as discrete photons, so even a perfect detector sees a slightly different number each second purely by chance. For the faintest targets this counting noise, not the optics, sets what is measurable, and the only remedy is more area or more time.
The current frontier is aperture again. Ground telescopes with mirrors close to forty meters across are under construction, built entirely from segments and dependent on adaptive optics to be worth their size. Their stated goal includes imaging small rocky planets around other stars directly. A handful of young giant planets, still glowing with the heat of their formation and orbiting far from their stars, have already been photographed. An Earth-sized world close in to its star is a far harder problem, and seeing one as light rather than as a dip in a curve would be a first.
Every Picture Is a Compromise
Every image of the distant universe on this site came through one of these machines, and every one of them is a compromise between area, sharpness, wavelength, and where you can physically put the thing. Knowing the compromises makes the pictures better rather than worse, because you can see what each instrument was built to catch.
It also puts the astonishing part in the right place. The remarkable thing is not that the images are beautiful. It is that a bucket of light, a limit set by the wave nature of light, and a mirror that reshapes itself a thousand times a second are between us and objects whose light left before Earth existed.
