Last Topic - Aberrations Lecture Index Next Topic - Atmosphere: Turbulence and Seeing

ASTR 5110, Majewski [FALL 2017]. Lecture Notes

ASTR 5110 (Majewski) Lecture Notes

Telescope Optics III: Modern Telescopes

Frontispiece from Bely, The Design and Construction of Large Telescopes.

In this webpage we explore topics relating to modern telescope development, including perspectives on the important drivers that specify telescope design, followed by a bit of a compendium of random topics relating to modern telescopes (domes and sites will be discussed in relation to seeing compensation).

Some References:

  • Chapters 1,4 of Bely.

A. Trends in Modern Telescope Development

Light Gathering Power

Clearly larger telescopes have increased light gathering power, and there has always been a push to build ever larger telescopes.

The overall trend in the development of the optical telescope has been a doubling of the aperture diameter about every 40 years:

From Bely, The Design and Construction of Large Telescopes.

This means, of course, a quadrupling of the light gathering power over the same timescale:

From Mountain & Gillett, "The Revolution in Telescope Aperture", Gemini Preprint #41.

However, the rate of increase in the light gathering power of the telescope only tells part of the story.

Other important trends that have influenced the true power of optical/IR telescopes are:

  • The quality and sensitivity of the instruments used with the telescopes, which affects the efficiency of the telescope.

  • The resolution achieved by the telescope -- which affects both the image quality and sensitivity (and thereby, again, the efficiency) of the telescope.

  • Useful field of view, which also affects telescope efficiency if one is interesting in covering large areas.

Instrument Sensitivity

Advances driven by:

  • Improvements in optical design, with substantial gains made by computer ray tracing and testing.

  • Improved optics (e.g., glass/substrates, coatings, fabrication).

  • New inventions (e.g., fiber optics, holographic gratings, filter fabrication, etc.).

  • Finding appropriate sites with improved transmission to certain wavelengths (especially ground-based infrared, spaced-based γ -ray, x-ray, UV, optical, IR).

  • Substantially improved detector efficiencies.


  • The invention of detectors (like the CCD) with quantum efficiencies (QEs) approaching 100% revolutionized modern astronomy.

  • For comparison, the photographic emulsion has a QE of only a few percent.

  • Thus, addition of a CCD camera to any telescope immediately increased its sensitivity/efficiency by a factor of ~30-40 over photographic use!! As if made the telescope diameters all ~5X larger!!

  • We will discuss detectors in more detail in a few weeks.

Resolution (and more Sensitivity)

From the ground, we typically never achieve the diffraction limit.

But the image quality is a strong driver in the power of a telescope, not only because of the better resolution offering more detail, but because more of the source power is concentrated into a smaller area on the focal plane, which improves the detectability of the source.

From Mountain & Gillett, "The Revolution in Telescope Aperture", Gemini Preprint #41.

Advances in telescope resolution (delivered image diameter) driven by:

  • Ability to quickly and automatically guide a telescope to decrease image blur due to tracking error and image wander.

  • An improved understanding of the origin and effects of seeing, and how to mitigate it (especially locally).

  • Locating observing sites (e.g., Chile, Hawaii) with more stable atmospheric conditions and therefore substantially better seeing --> ~2X improved resolution/sensitivity.

  • Improvements in mirror and dome design, which reduced local sources of seeing --> another ~2x improved resolution/sensitivity.

  • Active mirror figure compensation.

  • Active atmospheric compensation (e.g., fast shuttering, tip-tilt, active, adaptive optics).

  • Space-based platforms (e.g., HST).

  • Interferometry.

From Bely, The Design and Construction of Large Telescopes.

We will discuss a number of these topics in coming weeks.


Modern systematic surveys of the sky --- defined as surveys where there are stored images allowing quantitative analysis --- have a long history starting with the invention of astronomically useful* photographic plates (* Astronomical photographic plates need to have low reciprocity failure.):

  • The Carte du Ciel was an international (mostly European) project begun in 1887 by the Paris Observatory to take advantage of the new photographic technology. The goal was to cover the sky with 22,000 photographic plates of 2 degree x 2 degree each for the primary goal of astrometry.

    The project stretched on for decades, sucked away many valuable resources, and was never finished (although it did publish catalogs).

  • The Selected Areas designed by Kapteyn in the first decade of the 20th century was a more modest attempt to explore the Milky Way (which was "the universe" at the time) using 206 evenly spaced directions in the sky (i.e., not a full-sky survey).

    Kapteyn attempted to get observatories all over the world to contribute photographic images, astrometry and spectroscopy of stars in these Selected Areas.

  • The Palomar Observatory Sky Survey (POSS) was a more successful endeavor using the 48-inch Palomar (Oschin) Schmidt telescope, which made photographic plates covering 6 degree x 6 degrees on the sky and covered the northern sky throughout the 1950s.

    It's goal was to identify interesting targets to be explored with the Hale 200-inch telescope.

    The Palomar (Oschin) 48-inch Schmidt telescope. From

  • In the 1970s, better, finer-grained photographic emulsions were developed, and were used to make the Southern Sky Survey using the U.K. Schmidt telescope (which is a virtual copy of the Palomar Schmidt).

  • A new northern survey, the Second Palomar Sky Survey, POSS-II , was conducted throughout the 1980s, to take advantage of both the finer-grained and infrared sensitive photographic emulsions available. The POSS-II thus had the equivalent of B, V and I images.

    The survey has been digitized as the Digitized Sky Survey.

In the past decade, partly motivated by the huge success of the Sloan Digital Sky Survey (SDSS) and the Two Micron All-Sky Survey (2MASS) , there has been a surge of interest in ground-based, sky survey capability with digital cameras.

  • Obviously, efficiently covering the sky requires a telescope with a large field of view.

  • A metric used to gauge the power of a telescope to survey the sky efficiently, is encompassed in the expression

    "A Ω", where A is the collecting area of the telescope (in meters2) and Ω is the field of view of the telescope (in deg2).

  • This is also called the etendue or throughput of the system.

Here is a comparison of the etendue of some telescopes built with large fields of view (from the LSST website):

Two plots of the relative survey power in A Ω for various existing and planned telescopes. From .
Of course, the ability to capitalize on the etendue delivered by an optical system derives from an ability to pave the focal plane with detectors, which is an ongoing and significant technological challenge.

The proposed focal plane for the Large Synoptic Survey Telescope, which will include about 200 CCD detectors of 4k x 4k pixels each. From
For example, the "DES" shown in the histograms above is the "Dark Energy Survey" which is using the Blanco 4-m telescope (shown with a separate bar in the histogram) -- but using a new 2.2 degree field of view prime focus camera, which significantly increases the effective etendue of the telescope.

B. Large Mirrors: Shapes, Materials and Types

Telescope f /ratio

Especially for primary, one of most difficult decisions to make in designing telescope:

  • Type of primary mirror/ease of construction

  • Rest of optical train

  • Telescope structure/size/control

  • Instrument optical design/size (e.g., spectrograph sizes scale with focal length)

  • Size of dome/building


  • Potentially the site of the telescope (based on building footprint)

  • Overall cost (strongly a function of telescope length).

Goal has been to make primaries as fast as possible.

  • Cost driver

  • Smaller telescope tube can be made stiffer

  • Smaller wind cross-section

  • Secondary mirror smaller

Penalty is the stronger tolerances on secondary alignment (precision of placement varies as inverse cube of f /ratio).

From Bely, The Design and Construction of Large Telescopes.

Mirror Substrate

Mirror substrate has to be selected based on a number of considerations:

  • Must be available in large quantities, castable in large sizes.

  • Long term (century timescales) stability of shape.

  • Be polishable to required precision and be able to be coated with reflective materials.

    • Glass-like materials generally polish better than metallic ones.

    • May not care as much for longer wavelengths where microroughness more tolerable (λ/D larger).

    Before the invention of the process for silvering glass, the only practical choice for making telescope mirrors was polished metal. The metal of choice was "speculum", a mixture of 2/3 copper, 1/3 tin with a little arsenic, and possibly other stuff, like lead, silver or zinc. Speculum mirrors were also used in all kinds of personal household mirrors. This shows a picture of the 1.2-m diameter primary from William Herschel's "40-foot telescope" (referring, of course, to its length). From Wikipedia.

  • Rigidity. Must:

    • Sustain cooling from molten state, grinding, polishing, handling, etc. during construction.

    • Withstand constant motion/variable gravitational axis and wind loading.

    • Either not deform due to self-weight (flexure), or be flexible enough for active figure adjustment without breaking.

    Denser substances tend to be more rigid.

    The rigidity of a mirror (approximated as a thin plate) to deflection also goes as (thickness)3 / (diameter)2.

    • Overall stiffness depends on material selected and thickness of mirror.

    • Pick stiffness of mirror depending on how much control (authority) you want to actively shape mirror.

    • Segmented mirrors are low-rigidity because, though each rigid, can be moved w.r.t. one another.

    • Space telescopes can be low-rigidity because of low-gravity.

    From Bely, The Design and Construction of Large Telescopes.

  • Have good thermal behavior.

    • Ground-based telescopes have constant thermal changes adapting to air.

    • Space-based telescopes change temperature with solar attitude, but are basically working in the cold of space at cryogenic temperatures.

    • For homogeneous mirror, bulk temperature changes generally affect focal length but not figure (unless back/front gradients).

      Means need for telescope focus adjustments with temperature changes.

      Sometimes regular enough to be "mapped".

    • These effects minimized with low coefficient of thermal expansion (CTE), and maintaining a homogeneous bulk temperature.

    • Alternatively, one can have a substance with high thermal conductivity and low density and specific heat -- will diffuse heat quickly and come into rapid thermal equilibrium.

    • If one is actively shaping mirror figure, the CTE not as important, but still helpful to have high thermal diffusivity so mirror comes to air temperature quickly (to prevent mirror seeing).

The primary substrates used today have a variety of properties that must be weighed depending on application:

From Bely, The Design and Construction of Large Telescopes.

Lightweighting Mirrors

Goal is to make mirrors lighter without sacrificing rigidity.

  • Most of the bending stiffness of a deformed sheet is in the compression on the top surface and tension on the bottom.

    From Bely, The Design and Construction of Large Telescopes.

  • Middle section contributes little, and much can be removed (the principle of the "I-beam").

    Leads to the honeycomb mirror structure, of which there are completely closed and somewhat open back designs (see figure above).

  • The LBT, MMT, Magellan mirrors are cast into the open back design.

    Views of one of the mirrors for the LBT, with the honeycomb structure evident. Right picture shows the back side. Photos from John Hill's webpage,

  • The Hubble Space Telescope Mirror was assembled by fusing different-shaped pieces of low-expansion glass together into a sandwich with a closed back:

    From Bely, The Design and Construction of Large Telescopes.

  • With lightweighted, ground-based mirrors can also take advantage of the flexibility of the mirror and design mirror supports that control the shape of the optical surface:

    From Mountain & Gillett, "The Revolution in Telescope Aperture", Gemini Preprint #41.

  • The trend is towards lighter mirrors for all telescopes (an obviously important design feature of space-bound mirrors!).
    From Bely, The Design and Construction of Large Telescopes.

Segmented Mirrors

For some of the same reasons we discussed for size limits to refracting telescope objectives, monolithic mirrors are inherently size limited:

  • Fabrication and aluminization facilities scale with diameter.

  • Difficult to achieve substrate homogeneity over large volumes. Mechanical and thermal stresses grow.

  • Handling, transport difficulties.

    An 18-ton borosilicate "honeycomb" mirror for the LBT making its way up the mountain in November 2003. The mirror and its all-steel transport box weighed 55 tons. The trip included 122 miles of interstate and state highway followed by a 29 mile road with hairpin turns up the mountain.

  • Space telescopes limited to about 4-m for rocket fairing diameters.

Presently, about 8-m seems to be the breaking point for making monolithic mirrors economically.

Segmented mirrors offer advantages:

  • Lower mass.

  • Shorter thermal time constants compared to monolithic mirror of same size.

  • Easier handling and aluminization.

  • Ability to "change out" segments (helpful during realuminization).

  • Unlimited aperture sizes.


  • All segments must be figured to be parts of one parent shape.

    "Off-axis" paraboloidal segments tricky and expensive.

  • All segments must be kept precisely and actively aligned despite changing gravity, thermal effects, wind, etc.

Segmentation geometry:

From Bely, The Design and Construction of Large Telescopes.

  • "Petals" / "keystone" -- radial/azimuthal segments.

  • Hexagons, put down in rings.

    Hexagons have advantage of needing only one kind of support type.

For non-spherical primaries, number of different segments needed grows with the number of rings, as does number of edge sensors needed to sense positions and actuators to do simple tip/tilt/piston control of segments:

From Bely, The Design and Construction of Large Telescopes.

First multiple mirror telescope was by Guido Horn-d'Arturo, built in Bologna from 1920 to 1954 ("with brief interruptions"):

Guido Horn-d'Arturo's 1.8-m multiple mirror telescope made of 61 aluminized mirrors each about 20 cm across.

  • 61 mirrors 20 cm across.

  • All segments spherical and same.

  • Because of difficulties of alignment, fixed zenith telescope on the first floor of the Specola tower, with light coming through a hole in the terrace at the top of the tower.

  • The prime focus plate holder moved to track the stars.

Other famous multiple-mirror telescopes:

  • The Multiple-Mirror Telescope (MMT):

    • A new experiment in multiple mirror design, dedicated in 1979, Arizona.

    • Originally six identical 1.8-m mirrors on a movable alt-az mount.

    • Was third largest telescope in world when finished (a 4.5-m).

    • In the late 1990s converted to a single 6.5-m Arizona Mirror Lab mirror.

    The MMT before (left) and after (right) conversion from a six-mirror, 4.5-m equivalent telescope to a single mirror, 6.5-m telescope. See

  • Hobby-Eberly Telescope (HET) and South Africa Large Telescope (SALT)

    • HET (Texas) commissioned 1997, SALT 2005.

    • 11-m spherical primary, but at any given time only part of it is used (effectively a 9.2-m).

    • Fixed altitude mount (at altitude 55 degrees for HET, but 53 degrees for SALT, to optimize viewing of the Magellanic Clouds), but can be lifted and dropped to different azimuths.

    • Tracking secondary, can follow objects for up to two hours.

    The HET (top; photographs from University of Texas at Austin. The SALT (bottom) showing the basic design of both telescopes (courtesy SALT).

  • Keck I and II (Hawaii)

    • 10-m, 36 segments.

    • Note: Three rings of mirrors, requires *six* different mirror shapes (see Table 4.7 above, and the figure to the right shows the six nominal positions for the different [repeated] shapes).

    • Beams from both telescopes can be combined for interferometry.

    • Adaptive optics implemented.

    • Design cloned for Spanish telescope in Canary Islands (the GCT).

  • And of course, the Large Binocular Telescope (LBT):

    • Twin 8.4-m mirrors = 11.8-m collecting area and 22.8-m diffraction limited primary.

    • f/1.14 primary mirrors and f/15 Gregorian secondary.

    • Currently world's largest optical/IR telescope on one mount.

  • Still yet another multiple mirror geometry is that proposed for the Giant Magellan Telescope (discussed below):

C. Tubes, Trusses and Baffling

Small ground-based telescopes generally built with stiff, cylindrical tubes. But:

  • Heavy -- mass and size can't be kept reasonable as aperture increases.

  • Under gravity, the flexure bends the tube in a way that decollimates the optics.

  • Presents a large wind cross-section.

  • Prevents air from flowing across and cooling mirror.

Left figure from Bely, The Design and Construction of Large Telescopes. Right panel showing a Serrurier truss telescope, from

The Serrurier Truss was invented by an American engineer for the Palomar 200-inch as an open structure based on 8 struts making 8 isosceles triangles connecting a square base to a circular end. Two sets of these structures are connected at the square "box" to make a telescope with two circular ends.

  • When vertical triangles deflect, the parallelogram of horizontal triangles constrains the tube ends to move in a parallel plane.

  • This ensures that all optical elements remain aligned (collimated) no matter the flexure.

  • Invented for the Palomar 200-inch.

  • Used in many mid-sized telescopes.

  • One connects the altitude or declination axis to the box.

  • But being used less for big telescopes because of mass inefficiency and use of active optics.

As telescope mirrors become larger, the primary mass grows faster than the secondary, so center of mass moving closer to the primary.

  • Makes Serrurier design harder to implement.

  • Heavier telescopes difficult to support with only two compression and two tension members and only 4-point attachment.

  • Natural frequency of long truss members approach those incited by wind.

  • With active optics available now, can focus main worry on reducing wind effects and resonant frequencies of elements.

  • "Multi-bay" structures built using engineering finite-element analysis.

    (Finite Element Analysis: A method of [now computerized] analysis used in situations that are difficult to model by standard engineering techniques. Generally used to study mechanical structures and aids in calculating stress, shear force, load and other factors. The finite element method operates on the assumption that any continuous function over a global domain can be approximated by a series of functions/differential equations operating over a finite number of small sub-domains. The series of functions are piecewise, continuous and will approach the exact solution as the number of sub-domains approaches infinity.)

  • Also, use low CTE substances, or using counterbalancing elements with different CTEs that cancel one another.

From Bely, The Design and Construction of Large Telescopes.

Another concern, even larger now that we generally use open truss telescope "tubes" is scattered light.

Thus, most telescopes include in them various types of baffling to block or suppress unwanted radiation paths.

  • Generally conical or cylindrical tubes enclosing parts of the beam.

  • Often include perpendicular vanes to force radiation to make multiple scatters to get to detector (unlikely, so greatly attenuated, paths).

    From Bely, The Design and Construction of Large Telescopes.

  • The geometry of baffles (shape, location, beveling) has to be optimized for each telescope, and must balance the degree of blocking against the amount of tolerable beam obscuration (if any).

  • A particular concern for Cassegrain focus, because in this case focal plane faces the sky (unlike Nasmyth, prime, Coude foci) so there are possible direct ray paths to the detector.

    An example of a type of baffling combination is shown below.

    From Bely, The Design and Construction of Large Telescopes.

    Notice that the demands for wide field imaging (i.e., large focal plane detector surface) are more severe and result in potentially larger amounts of blocked light.

  • Other types of scattering possible within an optical system involve reflections off of the primary or secondary.

    The dominant source of this stray light in this situation is actually from scattering of off-axis rays off of dust.

    From Bely, The Design and Construction of Large Telescopes.

    From Bely, The Design and Construction of Large Telescopes.

  • Stray light for a space satellite, like Hubble, critically important, because observations are made in the presence of light from Sun, Moon and Earth.

    E.g., the Hubble Space Telescope baffling is extensive, because of Cassegrain arrangement.

    Includes numerous vanes and both secondary and primary conical baffles (all black).

    From Bely, The Design and Construction of Large Telescopes.

D. Mounts

Before 1980, nearly all telescopes were mounted with an equatorial mount:

  • Counteract Earth rotation by motion only on one, polar axis.

  • Simple correction with single speed.

  • No field rotation in focal plane.

From Bely, The Design and Construction of Large Telescopes.

A variety of implementations of the polar/declination axis combination:

The McCormick 26-inch is built with a German equatorial mounting.

From Bely, The Design and Construction of Large Telescopes.

Variation of a yoke equatorial-mounted telescope.
Now most telescopes are built with altitude-azimuth (alt-az) mounts.

  • Neither axis changes direction with respect to gravity.

  • Structurally sturdier than equatorial.

  • Less massive, less expensive.

Examples of alt-az mounted telescopes, both amateur and professional (WIYN 3.5-m). From and


  • Three axes of rotation needed: altitude (h), azimuth (A), and field rotation.

  • All three axes move with variable speed.

  • Could only do this with fast computers.

Conversion from RA-Dec (α, δ) to Alt-Az-Rotation (h, A, q):

From Bely, The Design and Construction of Large Telescopes.

From Bely, The Design and Construction of Large Telescopes. Note that the figure caption INCORRECTLY identifies the angle z as the "zenith angle", when it is in fact the ZENITH DISTANCE. The zenith angle is actually the angle marked A in the figure, and, for the particular configuration shown, is 360o-azimuth.

Field rotation is defined by the parallactic angle, q, which is the position angle (measured north through east at the target) of the arc connecting target to zenith.

The conversion from Right Ascension and Declination to Altitude, Azimuth and parallactic angle coordinates is given by:

where HA is the hour angle and φ is the observatory latitude.

E. The Biggest

Existing Telescopes

The largest telescopes in the world today, by hemisphere. Note that the teleascope labeled "ORM" (for "Observatorio del Roque de los Muchachos" on the Canary Islands) is a "copy" of the Keck Telescopes and is now named the "Gran Telescopio Canarias". From Click here and here for more outdated versions of this kind of comparison, made by the Keck consortium and the University of Hawaii.

More information tabulated about the world's largest optical/IR telescopes.

From Bely, The Design and Construction of Large Telescopes.

Some Proposed/Planned Large, Ground-based Telescopes

  • Large Synoptic Survey Telescope (LSST)

    • A planned 8.4-m telescope with enormous 10 square degree field to be built in Chile (Cerro Pachon, near the Genmini South Telescope and CTIO).

    • 3 billion pixel camera with a 9.6 square degree field.

    • Build up a deep survey image of the sky in multiple wavelengths (6 bands).

    • Will cover the entire southern sky with 2x15 second integrations every few nights.

    • Each piece of sky will be visited about 825 times in ten years.

    • 15 TBytes of date collected per night.

    • Goal is to find fast moving or variable objects.

  • Giant Magellan Telescope

    • Seven 8.4-m Arizona Mirror Lab borosilicate honeycomb mirrors (as of Fall 2017, four of the mirrors have been cast, the fifth will soon be started, and the first one made is heading to Chile).

    • Light gathering power equivalent to a 21.4-m filled aperture.

    • Diffraction limited resolution equivalent to a 24.5-m filled aperture.

    • f/8.4 Gregorian with adaptive optics secondary.

    • For Chile, at Las Campanas Observatory.

    • Partners: Carnegie Institution for Science, Harvard, Smithsonian Institution, Texas A&M, U.Texas (Austin), U.Arizona, Australian National University, Astronomy Australia Ltd., Korea Astronomy and Space Science Institute, University of Chicago, Sao Paulo Research Foundation.

  • Thirty Meter Telescope (TMT) project (for Mauna Kea, Hawaii):

    • 492-segment mirror, Ritchey-Chretien design.

    • A joining of several, previously separate efforts:

      • Formerly the California Extremely Large Telescope (CELT) -- Caltech/UC

      • Giant Segmented Mirror Telescope (GSMT) -- AURA

      • Very Large Optical Telescope (VLOT) -- Canada

    • Preferred site is Mauna Kea, Hawaii. However, shoudl that site become politically untenable, an agreement has been signed to host the telescope at Observatorio del Roque de los Muchachos on La Palma, Canary Islands, Spain.

    • Current partners: Association of Canadian Universities for Research in Astronomy, Caltech, University of California, National Astronomical Observatory of Japan, National Astronomical Observatories of the Chinese Academy of Sciences, Department of Science and Technology of India.

      (Left) Design of CELT. (Middle) Proposed TMT with car (red) and Hale 200" for scale. (Right) TMT design with all vents open to improve air flow and reduce mirror seeing.
    • The European Extremely Large Telescope (E-ELT) is a plan for a 42-m.

      • Replaces plan to build the Overwhelmingly Large Telescope, a 100-m telescope, which was deemed too complex and too expensive to build right now.

      • Proposed to have a "40-metre-class main mirror".

      • Even the E-ELT is estimated at nearly $1B cost.

      • 1000 elements to make up primary.

      • 5000 actuators.

      A comparison of the EELT to the VLT in size. Image credit: ESO.

    A new image like the above ones, including future telezcopes. From
    Finally, a comment on the scientific productivity of telescopes.

    Last Topic - Aberrations Lecture Index Next Topic - Atmosphere: Turbulence and Seeing

    Unless otherwise attributed, material copyright © 2005,2007,2009,2011,2013,2015,2017 Steven R. Majewski. All rights reserved. These notes are intended for the private, noncommercial use of students enrolled in Astronomy 511 at the University of Virginia.