An enormous mirror can be built on a mountain. Why fold a telescope into a rocket and send it somewhere much harder to repair?
Space telescopes rise above the atmosphere to collect light it absorbs, avoid its image distortion and reduce atmospheric thermal background. Which advantage matters most depends on the observation. Infrared telescopes also need carefully engineered cooling.
A larger mirror cannot restore missing light
Imagine observing the same star in visible light and X-rays. Much visible light passes through the atmosphere. Cosmic X-rays hardly reach the surface. A larger ground-based mirror cannot collect photons that the air has already absorbed.
Parts of the spectrum that pass readily to the surface are called atmospheric windows. Visible light, part of the radio range and selected infrared bands provide such windows. Elsewhere, molecules intercept the signal. NASA explains how water vapour, carbon dioxide and ozone cause this selective absorption.
So “infrared cannot be seen from Earth” is too sweeping. Some infrared reaches the ground, particularly at dry mountain sites; large portions remain blocked. For X-rays and many ultraviolet observations, instruments must rise above the absorbing layers.
Air also distorts and glows
Even transmitted light can arrive with a distorted wavefront. Moving air layers differ in temperature and refractive index. Light takes slightly different paths through them, making a star’s image tremble and spread. This is atmospheric turbulence, rather than absorption. Above the atmosphere, Hubble avoids this blur, as explained in ESA’s observatory FAQ.
A third obstacle is the warm atmosphere’s own infrared emission. A faint celestial signal has to be distinguished from this background. Mirrors and instruments emit too: going to space does not automatically cool them sufficiently. NASA identifies opacity, blur and bright infrared background as reasons for placing Webb in space.
Compare the fate of the signal in the diagram. Missing light, distorted light and extra background call for different solutions.
Three different obstacles between source and detector
- Signal from spaceX-raysWhat the atmosphere doesAbsorbsEffect on a ground-based instrument
The cosmic signal hardly reaches the surface.
- Signal from spaceVisible lightWhat the atmosphere doesMostly transmits, but distortsEffect on a ground-based instrument
Light arrives; turbulence blurs fine detail.
- Signal from spaceInfrared lightWhat the atmosphere doesTransmits selectively and emits its own radiationEffect on a ground-based instrument
Some windows are open; elsewhere the signal is absorbed or obscured by background.
Branch · engineeringOptional · you can skip this
Why doesn't adaptive optics replace space telescopes?
Adaptive optics measures wavefront distortion using a reference star or laser beacon and rapidly reshapes a mirror. Under suitable conditions, it can correct atmospheric blur extremely well. This helps ground observatories remain powerful partners to space telescopes.
It does not recover photons absorbed by molecules or remove all thermal background. Performance depends on wavelength, conditions and system design. “Ground telescopes are always less sharp” is as misleading as “adaptive optics makes launches unnecessary.”
Hubble and Webb have different eyes
Hubble is particularly useful in ultraviolet and visible light, while also covering some near-infrared wavelengths. Webb works primarily in the infrared. Their capabilities overlap, but Webb does not replace Hubble’s ultraviolet vision. NASA compares them as complementary observatories.
The distinction has a physical purpose. Visible and ultraviolet light help study hot stars and gas. Infrared reveals other features: emission from cooler objects, star formation behind some dusty veils and distant galaxy light stretched by cosmic expansion. We explored that last effect in the article on looking back in time.
Their locations differ too. Hubble circles Earth in low orbit. Webb travels around the Sun with Earth and follows a halo orbit near the Sun–Earth L2 point, roughly 1.5 million km from our planet. It neither hangs motionless at the point nor hides in Earth’s shadow. This orbit keeps the Sun, Earth and Moon on the same side of its sunshield.
Branch · quantitative comparisonOptional · you can skip this
What do the wavelengths and temperatures mean?
NASA’s overview gives Hubble an approximate combined range of 0.1–2.5 µm, and Webb 0.6–28 µm. A micrometre, µm, is one millionth of a metre. These are rounded limits for different instruments together, not a promise of equal sensitivity at every wavelength. Their overlap at 0.6–2.5 µm helps show why one is not simply a newer version of the other.
A multilayer shield protects Webb’s cold side to keep its own heat from overwhelming faint infrared signals. A cryocooler further cools the MIRI mid-infrared instrument’s detectors to below 7 K, approximately −266 °C. This is the temperature of particular detectors, not the whole observatory: other parts operate warmer. NASA describes the engineering and its purpose on its cryocooler page.
Choose a telescope for the signal
Predict this: could perfect adaptive optics let a ground telescope see cosmic X-rays through the atmosphere?
No. It corrects the wavefront of light that arrives; it cannot undo absorption. The instrument needs access above the atmosphere. For a faint mid-infrared source there is a further requirement: a sufficiently cold instrument and low background.
A launch opens particular wavelengths and cleaner observing conditions. A ground observatory can instead offer a large mirror and accessible maintenance. Start by asking what light needs measuring and what stops it reaching the detector. When you next see paired Hubble and Webb images, check their filters: differences can reveal the object itself, beyond the quality of the camera.