How JWST works

Why Webb works in the infrared

Visible light is a thin slice of the spectrum; infrared is the band of longer wavelengths, just past what the eye can see. Webb covers roughly 0.6 to 28 micrometres — from the red edge of visible light, through the near-infrared, into the mid-infrared. Three problems make that the right choice.

First, the universe expands, and expansion stretches light on its way to us: ultraviolet and visible light from the earliest stars and galaxies arrives redshifted into the infrared. Webb sees the distant past not because infrared is “older,” but because cosmic expansion has moved that ancient light into its band. Second, infrared passes through many clouds of interstellar dust that block visible light, so Webb can see into the nurseries where stars and planets form. Third, cold things — planets, dust, comets — are faint in visible light but glow in the infrared. What makes Webb powerful in the infrared is also what makes it demanding: the telescope must be kept extraordinarily cold, or its own heat would drown the signal.

The segmented mirror

Webb’s most recognisable part is its golden primary mirror: about 6.5 metres across, roughly 25 square metres of collecting area, built from 18 hexagonal segments. It is segmented because a single mirror that size would not have fitted inside the rocket; it launched folded and opened out in space. The segments are beryllium — light, stiff, and stable at very low temperature — under a whisper-thin coat of gold, which reflects infrared well (the whole mirror uses only a few tens of grams of it).

A big mirror does not “magnify” so much as gather photons: the larger the area, the fainter the object it can reach, and the faster it builds a usable signal. After launch the 18 segments had to be aligned to act as one surface, each nudged by tiny actuators until the errors shrank to tens of nanometres — a small fraction of the wavelength of the light they collect.

Kept colder than almost anything

Infrared light is heat. A warm telescope glows in exactly the band Webb is trying to observe — like watching a distant firefly with a lamp shining into the camera. So Webb is split into a hot side, facing the Sun, Earth and Moon, and a cold side carrying the mirror and instruments, separated by a five-layer sunshield the size of a tennis court. Each layer reflects and re-radiates heat sideways, so only a tiny fraction reaches the far side. The cold side sits near 40 kelvin (about −233 °C); the mid-infrared instrument, MIRI, must be colder still and carries its own refrigerator down to about 7 kelvin.

This is also why Webb orbits where it does. Rather than circling Earth like Hubble, it loops around the Sun–Earth L2 point, about 1.5 million kilometres out on the night side, where the Sun, Earth and Moon stay lumped on one side and the sunshield can hold them all at its back.

One observatory, four instruments

“JWST” names the whole observatory. The science is done by four instruments mounted behind the mirror, and it is worth keeping them distinct from the telescope itself:

• NIRCam, the near-infrared camera (about 0.6–5 µm), for deep images — and for aligning the mirror;
• NIRSpec, the near-infrared spectrograph (about 0.6–5.3 µm), which can record up to around 100 spectra at once through a grid of roughly 248,000 tiny shutters — an idea not far from DESI’s robotic fibres;
• MIRI, the mid-infrared instrument (about 5–28 µm), for cool dust, molecules, and the thermal glow of planets;
• FGS/NIRISS, which keeps the telescope locked on target and adds near-infrared imaging and slitless spectroscopy, including of exoplanet transits.

When a result is reported “from JWST,” it came through one of these. The instrument and the wavelength range matter as much as the telescope’s name.

From light to spectrum

A camera records how much light arrives; a spectrograph records how that light is spread across wavelength. Split a beam into its wavelengths and the smooth curve breaks into structure — bright emission lines, dark absorption lines, broad molecular bands — each the fingerprint of a particular atom or molecule. From a spectrum, astronomers can read what something is made of, how hot it is, how fast it moves, its redshift, and which molecules are present — things a plain image cannot show. How finely a spectrograph can separate neighbouring features is its resolving power, the wavelength divided by the smallest wavelength difference it can tell apart; higher resolution reveals more, but demands more light.

How Webb reads an atmosphere

Webb studies many exoplanets during transits — when a planet crosses in front of its star from our line of sight. As it does, a sliver of starlight grazes through the planet’s atmosphere before reaching us, and the gases there absorb their own particular wavelengths. By comparing the star’s spectrum during the transit with the spectrum just outside it, astronomers extract a transmission spectrum: the faint imprint of the atmosphere, printed onto the starlight. The depth of the dip also gives the planet’s size relative to its star — close to the square of the planet-to-star radius ratio.

This is delicate work. The atmospheric signal is a tiny fraction of the starlight, and it must be separated from the star’s own activity, detector noise, and instrument effects. Different instruments and wavelength ranges are sensitive to different molecules, which is why the same planet is often observed more than once, with more than one instrument, before anyone trusts a feature.

What Webb does not do

Webb does not photograph the Big Bang, see the whole sky at once, or deliver ready-made true-colour photographs — its light is largely invisible, and colour is assigned afterwards to real maps of intensity. It cannot punch through unlimited dust, measure a distance just by taking a picture, or identify life from a single molecule. It cannot even point wherever it likes at any moment: the sunshield must stay between its instruments and the Sun, so each target has its seasons. Every Webb result rests on sensitivity, exposure time, calibration, a physical model, and a careful statistical analysis — not on the picture alone.

In one sentence

Webb gathers faint red and infrared light with a cold, 6.5-metre segmented mirror, turns those photons into images and spectra with four instruments, and lets astronomers reconstruct the makeup of galaxies, stars, planets and atmospheres — one carefully measured spectrum at a time.