Atmospheric Effects - Seeing

Relevant syllabus point:

  • discuss the problems associated with ground-based astronomy in terms of resolution and absorption of radiation and atmospheric distortion

Our atmosphere is constantly in motion. It is a mixture of gases, water vapour, dust and other suspended particles. All these impact on the ability of a telescope to receive light and to clearly resolve an image.


The most obvious effect is that of absorption, most radiation incident on the Earth’s upper atmosphere does not reach the ground. The atmosphere is effectively opaque to all but some radio wavebands and light in the optical window. This includes the entire visible region (390 nm – 780 nm) plus the near ultraviolet, near infrared and some far infrared wavebands. Much of the infrared (21 mm to 1 mm) suffers absorption from water and carbon dioxide molecules whilst most ultraviolet undergoes absorption by ozone. Gas atoms and molecules absorb X-rays and γ-rays .


Scattering of light is strongest when the wavelength of the light is of the same order of magnitude as the diameter of the scattering particles. Visible light, therefore, is more readily scattered by dust and mist than infrared. This can be demonstrated by the use of a torch and an infrared remote control; visible light, from a bright, narrow torch beam for example, is readily scattered by fog or a cloud of fine particles such as talcum power or chalk dust. When shining a torch beam in or car headlights in fog, you may be dazzled by the back-scattered light. In contrast, infrared red radiation can pass through dust clouds readily. Aiming an infrared remote control through the cloud should still enable the settings on a television or video to be controlled. Replacing the dust cloud with a beaker of water (or a suspended clear plastic bag of water to avoid disputes about the glass) should see the infrared remote control ineffective due to the absorption by the water. The light on the other hand should pass through unaffected.


Variations in density of the atmosphere in a line of sight with an object cause intensity fluctuations. The variations in the refractive index of a cell of air above a telescope will alter the apparent position of an object, normally over a range of a few arcseconds. Collectively these effects combine to make point sources such as stars appear to “dance” about. This rapid change in brightness and position is termed scintillation. You may have used this to distinguish between a star and a planet merely by observing the night sky. Stars “twinkle” more than the bright planets because they are so much more distant that they are effectively a point source. Their light gets smeared across a seeing disc a few arcseconds in diameter. Planets, with an angular diameter of ~10-30 arcseconds are extended sources and thus less affected. Scintillation effects are worse for stars near the horizon where refraction effects are greater, leading to dispersion of light.

Comparison of two nights' seeing for the galaxy M 74
Credit: B. Keel, University of Alabama

Turbulent cells of air in the atmosphere limit the actual resolution of telescopes in the optical wavebands to typically no better than that obtained by a 20-cm telescope. The seeing at any location depends upon many factors and changes due to temperature, weather, pollutants and local microclimate. Good locations may achieve a typical seeing of 1 arcsecond but most sites are worse than this. Even the best sites rarely achieve seeing of better than 0.5 arcseconds. If the resolution a telescope achieves is limited by the seeing rather than its diffraction limit it is said to be seeing limited.

The images below simulate the seeing at two locations. The one on the left represents a telescopic image of a star from a telescope on Mt Fuji in Japan. The right hand image is from a telescope in the Andes in Chile.

Amimated image of 'seeing' on Mt Fuji
Mt Fuji, Japan
(Poor seeing)
Animated image of 'seeing' in Chile
Andes, Chile
(Good seeing)
(Images adapted from CERES MountainQuest! activity)

Where to Build Telescopes?

Modern large ground-based optical, infrared and radio telescopes are complex and expensive pieces of research equipment. A crucial part of their effectiveness is determined by their location. Even with the advent of powerful new techniques such as adaptive and active optics and interferometry to improve the resolution of images, the full potential of instruments can only be fulfilled if the telescopes are built at the best possible location. So what are the criteria? Consider the locations for some large optical telescopes on the diagram below:

Locations of some large optical telescopes

Optical and Infrared Telescopes

Modern instruments are built in only a handful of places, notably Mauna Kea in Hawaii and the Atacama Desert region of the Chilean Andes. Other large telescopes have also been built on mountains on the west coast of California and on the Canary Islands in the Atlantic Ocean. These sites share some common factors. They are at high altitude therefore above a lot of the local weather and atmosphere. They are also either on the west coast of continents or on mid-ocean islands. This is no coincidence. Given that the typical weather patterns move from west to east it means that the air masses moving over the telescopes have traversed a broad expanse of ocean and so have not picked up much in the way of pollutants. As the air rises as it hits the coast and mountain it tends to lose any moisture low down. As it has not had time to be heated by large expanses of warm ground it does not become too turbulent. The result is a laminar flow of air across the observatory. The lack of turbulence provides sub-arcsecond seeing whilst the absence of water vapour (in part due to their high attitude) makes these sites ideal for infrared and millimetre wave observations. A complete list of large optical telescope with links can be found at the SEDS site, Large Telescopes.

Australia unfortunately does not possess any site that fits the above criteria. Our largest optical telescopes are at Siding Spring, in the Warrumbungles mountains near Coonabarabran in NSW. The largest of these is the 3.9m Anglo-Australian Telescope (AAT), opened in 1974. The median seeing for the AAT is about 1.2 - 1.3 arcseconds with about 65 - 70% usable nights per year. Recent air conditioning of the AAT dome will hopefully see the median seeing reduced to 1 arcsecond. By comparison, the seeing at Mauna Kea is typically about 0.8 arcseconds and can get down to 0.4 arcseconds on exceptional nights. This is one reason why Australia joined the international Gemini consortium. This give Australian astronomers access to the two 8.1m Gemini telescopes, one in the northern hemisphere on Mauna Kea, the southern hemisphere one on Cerro Pachón in Chile. These identical telescopes are optimised for visible and near infrared observations and utilise adaptive and active optics to provide high resolution images.

Schematic showing the global coverage of the Gemini telescopes. (Image: Gemini Observatory)

One site that has excellent characteristics for infrared ground-based astronomy is high on the Antarctic plateau. The altitude and lack of water vapour in the dry environment have the potential to provide an excellent site that could allow continual observation of some objects during the long antarctic night. Australian astronomers have been active in investigating the potential of Antarctic astronomy. For more details, visit JACARA, the Joint Australian Centre for Astrophysical Research in Antarctica.

Radio Telescopes

Many radio telescopes traditionally have been located in low valleys where the surrounding hills help block extraneous radio emissions from terrestrial sources. This radio frequency interference (RFI) is an increasing problem for radio astronomers due to the spread of mobile telephony devices, digital television transmission and the like. It is analogous to the problem of light pollution faced by optical astronomers. Fortunately Australia is relatively 'radio-quiet' due to our low density of population. This is one reason why future generation radio telescopes such the SKA (Square-Kilometre Array) may be built in western Australia.

Whilst weather is less of a concern and observations can take place 24 hours a day, large dishes such as the Parkes radio telescope are sometimes affected by strong winds. As radio astronomers have moved to observing at higher frequencies in the mm-wavebands, absorption due to water vapour in the atmosphere has become a problem too. Some mm and sub-mm observatories are therefore being built at high altitude in similar locations to large optical telescopes. Examples include ALMA (Atacama Large Millimeter Array), SEST (Swedish-ESO Submillimetre Telescope) which are both in Chile and the SMA (Submillimeter Array) and JCMT (James Clerk Maxwell Telescope), both on Mauna Kea, Hawaii.

Telescopes in Space

Astronomers wishing to observe at other wavebands have little option but to place telescopes in space. Most people are familiar with the wonderful images produced by the HST (Hubble Space Telescope), a 2.4m optical, ultraviolet and near IR that orbits 600km above the Earth. Apart from its wide observational waveband (120 - 1000nm) one of its key features is its angular resolution of about 0.08 arcseconds This is equivalent to being able to separate two flies 1m apart in Perth when viewed from Brisbane.

The image below shows a comparison of images form the 8m Subaru telescope on Mauna Kea and the 2.4m HST using the ACS (Advanced Camera for Surveys) of a region of sky. The Subaru image was taken when seeing was about 0.8 arcseconds though it can get down to 0.4 arcseconds under exceptional conditions on Mauna Kea. The resolution for the HST image is 0.8 arcseconds

Comparison of resiolution from Subaru (ground-based) & HST (space-based)
Image: NASA, Mauro Giavalisco, Lexi Moustakas, Peter Capak, Len Cowie and the GOODS Team.

The resolution advantage of HST over ground-based telescopes will decrease as adaptive optics systems become more widespread on 8m class telescopes. One advantage of the larger ground-based telescopes is their sensitivity as their primary mirrors are much larger. They also do not require adaptive optics to correct for atmospheric effects.

Space-based telescopes are essential for most infrared, ultraviolet, X-ray and γ-ray observations. Developments over the last four decades have seen such telescopes improve in resolution, sensitivity and operational life spans. Space-based astronomy however still faces many difficulties not faced by telescopes on Earth. Problems include:

  1. Cost. It is much more expensive to design, build, launch and operate a telescope in space. As a comparison, the HST cost about US$2 billion to build and launch. Unique among space-based telescopes in being designed for regular upgrades, the HST has had four service missions. These visits by astronauts on Space Shuttle flights provide maintenance and install new instruments. A single service mission may cost as much as US$700m. The Gemini project of two 8.1m telescopes cost US$184m. The VLT project in Chile cost $600m for 4 x 8.2m telescopes with additional 1.8m instruments. Once the telescopes are linked together to form an optical interferometer, the VLTI, it will have a collecting area equivalent to a 16m dish and an effective aperture of 200m, giving it an angular resolution of 4 milliarcseconds, 25 times sharper than the HST.
  2. Lifespan. The lifetime of most space telescopes is limited by the amount of onboard fuel they can carry for corrections and orbital adjustment. They also normally rely on gyroscopes for control and pointing. If these fail, it becomes difficult or impossible to control the telescope. Infrared telescopes and some others rely on cryogenic gases to cool sensors. Once the onboard supply is used up, the mission may have to end.
  3. Risk. It is inherently risky launching satellites into space. Some space astronomy missions have failed to catastrophic failure of the launch vehicle. Others, such as Hipparcos have had to be curtailed or modified due to incorrect orbit insertion. Once in space there is the chance of damage from collisions from orbital debris or meteors. Solar flares and cosmic rays also pose a problem, especially for instruments beyond the protective Van Allen belts.
  4. Size. It is extremely expensive to launch objects into space. Using the Space Shuttle it is about US$20000 per kilogram. Apart form the cost another constraint is the restrictions on the size of satellites. The primary mirror for the HST was limited in size as it had to fit into the cargo bay of a Shuttle. HST's planned successor, the James Webb Space Telescope will use a 6.5m segmented mirror made of beryllium that will be folded up for launch on an Ariane 5 launch vehicle. The mirror will be only one-third the mass of Hubble's but will require active optics to maintain its correct shape. As the JWST will be placed at the L2 point 1.5 million km from Earth it will not be able to be maintained or upgraded by visits from astronauts. The size constraints pose limits on the apertures of mirrors with subsequent impact on sensitivity and resolution.
  5. Upgrades and Maintenance Ground-based telescopes can be upgraded relatively easily. Telescopes such as the AAT, Gemini and the VLT are improved throughout their lives by the addition of new instruments and sometimes improved optics. Mechanical and software problems can be fixed by engineers and scientists. Apart from HST, space telescopes can not have new instruments added of mechanical problems fixed after launch. Some can have bug fixes and new software uploaded to improve their operation but even this can be tricky.

Further Information

Large Telescopes is a regularly updated site tabulating all the large optical telescopes in the world. It is organised according to operational status of telescopes and lists them by aperture, providing additional information on name, location, latitude and longitude and brief comments. Most telescopes are linked to their relevant homepages. This site is a valuable source of data and a good reference for students to commence any secondary research task.

Gemini Observatory is the public home page for the Gemini telescope project. Links are provided to a range of pages including those dealing with the engineering and technology, observations, media releases and images. The Australian Site discusses Australia's role in Gemini and why the telescopes were not located in Australia.

M74—the Effects of Astronomical Seeing provides a useful split-frame image of the galaxy M74 that shows the effect that atmospheric turbulence has in the resolution obtainable from ground-based observations.

The Purpose of a Telescope is a simple page that addresses seeing, resolution and diffraction limits for telescopes, linking to other pages in this set of lecture notes. The page on seeing has a nice set of four images showing the effects of seeing on an image of galaxy clusters.


  1. Describe two reasons why stars near the horizon appear to "twinkle" more than stars high overhead.
  2. Why is high altitude important in selecting a site for an infrared telescope?
  3. Why is cooling the interior of a telescope dome during the day an effective way of improving the seeing?
  4. The Australia Telescope Compact Array is used for mm-wave observations during the winter months but not in summer. Suggest a reason for this.
  5. Construct a table comparing the cases for a 4m class optical/near infrared telescope in a) Snowy Mountains of NSW, b) Atacama desert in Chile, c) Antarctic Plateau, d) Orbiting Earth. In your table you should clearly identify the benefits and problems with each case in a concise form.

La Palma GTC 10.4m telescope South African Large Telescope AAT VLT Gemini Observatory Gemini Observatory Mauna Kea telescopes HET LBT Kitt Peak National Observatory Palomar Observatory Hale Telescope Keck telescopes Subaru Telescope