Telescopes and instruments

A telescope does two things: collect photons and resolve directions. Aperture governs both, which is why the history of astronomy is largely a history of larger mirrors.

Collecting area and resolution

Sensitivity scales with collecting area, so with $D^2$. Angular resolution scales with $\lambda/D$. Doubling a mirror’s diameter therefore gathers four times the light and halves the resolvable angle, which is why aperture is the single most quoted number about any telescope.

Note that a ground telescope only achieves these figures with correction. See Space telescopes for why some wavelengths require orbit regardless.

Optical design

Refractors use lenses and are limited by chromatic aberration and by the practical impossibility of supporting a large lens, which can only be held at its edge. The largest ever built, the Yerkes 1.02 m refractor of 1897, remains the largest and will stay so.

All large telescopes are reflectors. A mirror can be supported across its back, has no chromatic aberration, and can be made from low-expansion materials.

Large mirrors are segmented rather than monolithic beyond about 8 m, because casting, transporting and supporting a larger single piece becomes impractical. Keck pioneered this with 36 segments; the ELT will use 798. Each segment needs active position control to nanometre precision so the array behaves as one surface.

Adaptive optics

Atmospheric turbulence limits ground-based resolution to about one arcsecond at a good site regardless of aperture. A 10 m telescope without correction has the resolution of a 15 cm one, and merely collects more light.

Adaptive optics measures the distorted wavefront hundreds of times per second and applies the inverse distortion with a deformable mirror. Performance depends strongly on wavelength: turbulence is far easier to correct in the infrared than the visible, because the coherence length scales as $\lambda^{6/5}$.

A reference source is needed, and natural stars bright enough are rare. Laser guide stars excite sodium atoms at about 90 km altitude to create an artificial reference anywhere in the sky. They cannot sense overall image motion, since the laser is deflected on both the way up and the way down, so a natural star is still required for tip-tilt correction.

The consequence is a substantial shift in the ground-space balance. A 39 m ground telescope with adaptive optics achieves better infrared resolution than any orbiting telescope, at a fraction of the cost, and can be upgraded. Space retains the advantage in wavelengths the atmosphere blocks, in stability, and in thermal background.

Detectors

The move from photographic plates, with a few per cent quantum efficiency, to CCDs above 90% was equivalent to enlarging every telescope in the world by a factor of about five in diameter, and it happened in a decade.

Instruments

Spectrographs disperse light to measure composition, temperature, velocity and redshift. The key parameter is resolving power $R = \lambda/\Delta\lambda$: a few hundred suffices for redshifts of faint galaxies, while measuring stellar radial velocities at metres per second requires $R$ above 100 000 and extreme wavelength calibration stability. See Exoplanets.

Coronagraphs suppress light from a bright source to reveal faint companions, using masks in the focal and pupil planes to control diffraction. Achieving the 10¹⁰ contrast needed for an Earth analogue at a fraction of an arcsecond is the central technical challenge of direct exoplanet imaging.

Polarimeters measure polarisation, which traces magnetic fields, scattering geometry and dust grain alignment.

Integral field units record a spectrum at every point in a two-dimensional field, producing a data cube rather than an image or a spectrum. This has become standard on major instruments because it removes the need to choose in advance where to place a slit.

Interferometry

Combining separated telescopes synthesises an aperture as large as the separation between them. Resolution becomes $\lambda/B$ with $B$ the baseline, which can vastly exceed any single mirror.

Interferometry is far easier at longer wavelengths, because the signals must be combined with a path length accuracy of a small fraction of a wavelength. At radio wavelengths the signals can be recorded with precise timestamps and correlated later, which is what makes intercontinental baselines possible. At optical wavelengths the light must be physically brought together with sub-micron path control, which limits baselines to the scale of a single observatory.

The cost of interferometry is that it does not produce an image directly. It samples the Fourier transform of the sky at spatial frequencies corresponding to each baseline, and an image must be reconstructed from incomplete sampling. That reconstruction involves choices, which is why the Event Horizon Telescope results were produced by independent teams using different algorithms as a cross-check. See Black holes.

References

• Schroeder, D. J. Astronomical Optics, 2nd ed., Academic Press, 2000.
• Hardy, J. W. Adaptive Optics for Astronomical Telescopes, Oxford University Press, 1998.
• Thompson, A. R., Moran, J. M. and Swenson, G. W. Interferometry and Synthesis in Radio Astronomy, 3rd ed., Springer, 2017. Open access.
• Rieke, G. H. Detection of Light: From the Ultraviolet to the Submillimeter, 2nd ed., Cambridge University Press, 2003.
• Event Horizon Telescope Collaboration, ApJL, 875:L4, 2019, on imaging and reconstruction methodology.