Site selection is critical for a telescope. As long as
observations are ground based, the observation performance is strongly
affected by the Earth's atmosphere. A resolution of a telescope is determined
by intensity of atmospheric disturbance. The background radiation in
infrared wavelength is strongly affected by the amount of water vapor and
air temperature. Although resolution can be improved through use
of an adaptive optics system, better performance will be obtained by ensuring
atmospheric disturbances are inherently weaker.
Conditions that should be satisfied by a site
Factors to be considered in selecting a telescope
site are summarized in Table shown below. Clear weather must be available for a
large proportion of night time, and seeing (stellar image diameter
in a long exposure) must be good. Beyond those factors, for a large telescope
in which adaptive optics will be frequently used, a height distribution
of layers in which atmospheric disturbance (including disturbance of
refractive index, related to temperature disturbance) occurs, a speed
of disturbance, and a spatial spectrum of the disturbance (particularly
in a low spatial frequency region) are also important. The height distribution
of the layers affects a width of a field of view in which adaptive
optics correction can be applied, while the speed of disturbance change
and the spectrum in a low frequency region have a considerable effect
on difficulty of designing an adaptive optics system ( a speed of
control and amount of correction required). A good site has weak disturbances
that are distributed mainly at low altitude, exhibit slow fluctuation,
and which have small low spatial frequency components. In addition, it
is important in telescope design that these conditions do not fluctuate
over time.
In a case of infrared observations, it is important
not only that a stellar image discussed above be clear, but also that
background light be dim and that there be little absorption by an
atmosphere. Principal components of background light are that emitted
from molecules containing OH at wavelengths up to 1.6μm, and thermal radiation
from the atmosphere itself at longer wavelengths. Consequently, a site
with relatively little overlying air (i.e., a high-altitude site) and low
air temperature is generally preferable. Atmospheric absorption is dominated
by absorption due to water vapor, the effects of which are relatively small
at high altitudes and low temperatures. Significant absorption in infrared
wavelength also occurs due to molecules such as CO
2, CO and O
3.
Sources of
background light in optical wavelength include, in addition to air glow,
artificial light from cities, and auroras in polar regions.
Conditions on the ground, as well as sky conditions, are important to telescopes
and observation instruments. A difficulty of designing a large telescope
is reduced somewhat if a wind at ground level is weak and a wind speed
and direction do not fluctuate to a considerable degree. It is also important
from the point of view of structural design that the site be unaffected
by strong earthquakes. Adjustment of instruments is also simplified if
diurnal and seasonal temperature variations are small.
In addition to the natural conditions discussed above,
human activity must be considered in construction. Even
if observations are conducted remotely, occasional access by staff will
be necessary for
maintenance and modification. In addition, safety and
possible future expansion depend on maintaining good relations with
local residents and government,
and on freedom from civil strife. It is also necessary
to collect information on a wide variety of local conditions such as whether
the site is in a
flight path (or will be in future) and whether there
are plans for mine development, which could, for example, raise dust and
bring urban development with accompanying pollution.
| Factors to be considered in selecting a telescope site： |
| proportion of clear weather at night |
dust, aerosol |
| seeing |
wind speed at ground |
| precipitation |
variation of wind direction and speed |
| altitude distribution of atmospheric disturbance |
earthquake |
| temporal spectrum of atmospheric disturbance |
variation of temperature |
| spatial spectrum of atmospheric disturbance |
infrastructure (road, electric power etc.) |
| infrared background |
cooperation with regional government and poeple |
| airglow, aurora |
flight route (in future) |
| visibility |
developement status/plan |
Methods of site evaluation
In evaluating a site, a large number of meteorological
statistics will become necessary. For this
purpose, historical artificial satellite data should
be used first to determine a proportion of clear weather,
upper level winds, and a distribution of water
vapor as a basis for site selection. However, uncertainties
exist in the process of converting satellite data to
meteorological quantities, and ground monitoring at candidate
sites remains essential.
Proportion of clear weather
Locations having high proportions of clear weather can
be selected using global meteorological
satellite data. Once candidate sites are selected, ground-based
visible and infrared cloud monitoring is necessary.
Precipitation
Precipitation data are generally obtained by conversion
from water vapor absorption near 1μm or from optical
thickness of the atmosphere at submillimeter wavelengths.
The latter method is more accurate when precipitation
is less than 1mm. These measurements can also be made using GPS.
Seeing, altitude distribution and spectrum of atmospheric disturbance
Seeing through all layers of the atmosphere is measured using a differential
image motion monitor (DIMM). Since turbulence near the
ground accounts for a large proportion of atmospheric
disturbances, it is desirable for DIMM measurements to
be carried out at the height at which the telescope will
be constructed. Even at the same location, a difference of height of only 2-3 m can
lead to a greatly different measurement result. Consequently,
when only DIMM measurements are used, comparison among
different sites is difficult. It is necessary to measure
an effect of the layer adjacent to the ground at the same time.
If height distribution of the disturbances can be
measured, comparison among different sites will become
more accurate. For this reason, observations should be
carried out with SCIDAR or MASS, which use star scintillation
as an indicator, or SODAR, which uses scattering of
sound waves. Another possibility is direct measurement using balloons.
As SCIDAR uses double stars as light sources, directions and time
of observations are limited. In addition,
a telescope with aperture on an order of 1m is needed
to measure a turbulent layer at altitudes of about 10km.
Hence, measurements with SCIDAR
are not easy. It should be noted
that a type of SCIDAR that uses single stars as light
sources has recently been proposed.
MASS is small and simple, making the method useful for
measurements at locations without good infrastructure,
but the sensitivity for low-altitude
turbulence is relatively low. For measurements of an atmospheric boundary layer,
which accounts for a large component of disturbance,
it is necessary to use SODAR together with
other instruments. As SODAR does not detect features of the lower air layer
below a height of about 30m, a weak heat turbulence meter (for example)
is installed on a tower to measure turbulence near the ground.
Due to the difficulty of measuring a turbulence spectrum,
it is usual to assume a shape for the spectrum
when measuring a coherence time and a wavefront outer scale,
which are important parameters affecting the design of an adaptive optics system.
The coherence time is the time scale of speed of fluctuations of atmospheric
disturbances, and is measured from magnitude of star scintillation.
The wavefront outer scale is the space scale on the low-frequency side
at which the Kolmogorov law of a power of 2/3 breaks down, and is estimated
from the movement of a star measured along several baseline
lengths. Further development in techniques for measuring the outer
scale is needed for evaluation of the turbulence spectrum.
It is also necessary to rigorously
calibrate measurement instruments used in these evaluations to ensure accurate
comparison among sites.
Locations suitable for construction of an astronomical
observatory
Mauna Kea
More than 10 telescopes are now in operation on Mauna
Kea, which is today one of the main centers of astronomical observation
in the world.
This is a typical isolated mountain on an oceanic island. The altitude
is greater than 4100m, which is well above the upper boundary of the temperature
inversion layer (ca. 2000m), meaning that there is little water vapor despite
being surrounded by ocean. The proportion of clear weather is less than
in Chile. The infrastructure is well developed, but on the peak of the
ridge, where observation conditions are best, there are already many telescopes,
and the remaining available land is limited. There are also restrictions
on the construction of new structures and modification to existing structures
for environmental protection reasons.
Chile
Central and northern Chile are located at latitudes
with a cold ocean current flowing along the coast, leaving
a large expanse of dry land. The ocean side of the Andes, including Cerro Tololo,
La Silla, Paranal and Atacama, is characterized by high altitudes and a
dry climate, which have made this area another major center for astronomical
observations, with sites comparable to Mauna Kea.
The Chajnantor area of Atacama has a broad expanse of flat plain at altitudes
near 5000m. Construction of Atacama Large Millimeter/Submillimeter
Array (ALMA) through cooperation among North American, European and Japanese
radio observatories has already begun. It is thus expected that this area
will have a well developed infrastructure. ALMA is being built on a flat
plain, but nearby there are peaks higher than 5500m. Among these is Cerro
Chajnantor (5700m), which the University of Tokyo is surveying as a possible
telescope construction site. Data on seeing are as yet sparse, but there
are indications that the seeing can be expected to be as good as that on
Mauna Kea or better. This site is near the Bolivian border, where a proportion
of clear weather is somewhat less than that on Cerro Paranal. Additional
surveys are necessary.
Southwestern United States and northern Mexico
This region is dry for reasons similar to those in Chile.
CELT group is searching for possible sites in this area as well as
on Mauna Kea and in Chile. UNAM, a Mexican group planning a 6.5m telescope,
is surveying San Pedro Martir as a candidate site.
La Palma
La Palma is located in the Canary Islands off the coast
of Morocco. It is an isolated mountain on an oceanic island, like Mauna
Kea. The peak altitude is 2400m, which although not high is above the upper
boundary of the temperature inversion layer (1500m), meaning there is relatively
little water vapor. At present, GTC (Gran Telescopio Canarias, a telescope
with 10m segmented primary mirror) is under construction at this site.
As the site is near the Sahara Desert, it is affected by dust, particularly
in summer, when dust can reach an altitude of 3000m.
Southeastern Uzbekistan
This location is near the Pamirs and the western end
of the Altai Mountains. This site has been based for astronomical observatories
since the 1960s. In the late 1990s, seeing was
reevaluated, and it became widely known that there are good sites even
in this interior continental area, which does not belong to either of the
above categories of site generally considered to offer good conditions.
The area is characterized by relatively slow changes in atmospheric disturbance
(time scales 2 - 4 times longer than at other principal sites), making this
area very favorable for adaptive optics.
Antarctica
The South Pole is very cold, and the air is very dry.
Infrared and submillimeter telescopes have now been operating
there for more than 10 years. However, as the seeing is poor, the location is not
suitable for large telescope requiring good spatial resolution.
France and Italy are jointly developing Dome C (75°
06' S, 123° 21' E) as a site, and an Australian group
is now cooperating in those efforts. The altitude
is 3280m, and an amount of water vapor is very low.
The Antarctic plateau is located in a region of air subsidence,
resulting in a high proportion of clear weather and little snowfall (ca.
10cm per year). Measurements of seeing in winter have been initiated.
Although data are as yet sparse, the preliminary indications are that seeing
is better than 0.3”, giving this site possibly the best observation
conditions in the world. In addition, disturbances primarily occur at low
altitude, resulting in a large isoplanatic angle and good coherence time.
Surveys are continuing. Although auroras can become a source of background
light in Antarctica, Dome C is sited near the present location of the South
Magnetic Pole where auroras are infrequent.
As a bedrock is not exposed at Dome C, all structures must be built on
top of the thick Icecap. In contrast to other areas, it is necessary to
deal with problems specific to Antarctica, including subsidence and movement
of structures. In addition, it is necessary to consider that only the southern
sky can be observed in Antarctica and that the sun does not set in summer,
requiring observation objectives to be limited accordingly. Dome C is accessible
by aircraft. It is planned that it will become possible to winter over
at Dome C from 2005.
Japan operates a station at Dome Fuji (77° 19' S, 39° 42' E), which sits
at an elevation 3810 m higher than Dome C. Scientists
have now wintered over there for 9 years. It is possible
that Dome Fuji offers conditions as good as
those at Dome C or better. However, as Dome Fuji is closer
to the aurora belt than Dome C, background light is expected
to be a problem, particularly in optical wavelength.
Other areas
China has begun a survey to find the best possible observation
site in the country. Attention is being focused on western
Tibet. Japan is cooperating in this survey. Western Tibet
is a high plateau, exceeding
4000m, and the amount of water vapor is relatively small.
There is a monsoon effect, but from satellite data the
proportion of clear weather is equal
to or better than 70% experienced at Hanle in India.
Although there is the problem of turbulence generated
by the Pamirs and Karakoram mountains,
there remains a possibility of finding good sites, and
surveys are expected to continue.
The Greenland Icecap can be expected to have weather conditions similar
to those in Antarctica. The altitude is 2000m in the north, and 3000m in
central Greenland. Toward the south, the chances of being affected by aurora
increase. This area has not yet been surveyed.
