Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.naic.edu/~tghosh/haat/smallT4dc_15sep08.doc
Дата изменения: Tue Sep 23 21:30:21 2008
Дата индексирования: Sun Apr 10 18:57:16 2016
Кодировка:

Поисковые слова: arp 220

A SMALL AUXILIARY ANTENNA AT ARECIBO
C.J. Salter and
Tapasi Ghosh


1. Introduction

The 305-m Arecibo radio telescope is the world's largest, most sensitive,
single-dish radio telescope. It is equipped with receivers between 47 MHz
and 10 GHz. In addition to its single-dish capabilities, the telescope also
participates in Very Long Baseline Interferometry (VLBI) observations with
the VLBA, HSA, EVN and Global VLBI networks.

In the quest for higher sensitivity, VLBI observations have been evolving
over the years. For a given VLBI array, this can be achieved by either
increasing the bandwidth or the integration time. While bandwidth
limitations come either from equipment (in the case of continuum immaging)
or from the natural phenomena under study (spectral lines), phase
fluctuations due to the propagation of the signal through the Earth's
troposphere and ionosphere limit the basic coherence time of VLBI
observations. The technique of "phase-referencing" in VLBI allows the
possibility of correcting for these effects and thereby increasing the
coherence time by large factors.

1.1 VLBI Phase Referencing:

Phase referencing in VLBI observations has made it possible to study very
weak radio sources by increasing the effective coherence time from, at
maximum, a few minutes to hours. Currently, some 50% of VLBI observations
are carried out using this technique.

Phase-referenced observations can be performed in two modes, nodding style,
and in-beam. In nodding-style phase referencing, the antennas switch
between the target source and a nearby calibrator, called the phase
reference, every few minutes. The duration of one cycle of observing the
target and phase reference is called the cycle time, and is typically about
5 minutes. This procedure can be successfully carried out for observations
at 1 GHz and above. However, at frequencies below 1 GHz the raw coherence
times become very short due to ionospheric effects, requiring the phase
calibrator to lie within the (voltage) primary beam of all antennas in the
array.

However, phase-referenced VLBI observations with the Arecibo 305-m
telescope encounter limitations since the Gregorian dome, located on a
suspended platform, has slow slew rates (24o/min in azimuth, 2o.4 /min in
zenith angle.) Hence, in a typical phase-referenced observation where the
calibrator could be located 3o or more from the target, a significant
amount of observing time, often ™ 50%, is wasted slewing between the two
sources, leading lower signal-to-noise ratios. However, phase-referenced
VLBI could be performed using a smaller "Auxiliary" Telescope (AT) to track
the phase calibrator, while the 305-m antenna observes the target most of
the time, only occasionally moving to the calibrator. The effects due to
ionospheric/tropospheric phase fluctuations can then be derived from the
small-telescope data and applied to the target data from the 305-m dish.
While this technique has been successfully applied for observations with
the 220-km baseline MERLIN array in the UK, it would be a new approach for
VLBI. In fact, implementing its VLBI application will be original research
in its own right.

2. Areas of Astronomical Research Benefiting from the Auxiliary Telescope
(AT) in Phase Referenced VLBI

2.1 Stellar (radio) Astrometry: In a white paper submitted to the NSF
ExoPlanet Task Force, Bower et al. (arXiv:astro-ph/0704.0238v1) explored
the possibility of "Radio Astrometric Detection and Characterization of
Extra-Solar Planets". Utilizing the better than 100-microarcsec positional
accuracy routinely achieved with the VLBA, they propose carrying out a
Radio Interferometric PLanet search (RIPL) that will survey 29 low-mass,
active (radio-loud) M-dwarf stars over 3 years. This would have sub-Jovian
planet mass sensitivity at distances of ~1 AU from the star. They note
that, "Radio astrometric planet searches occupy a unique volume in planet
discovery and characterization parameter space, which gives greater
sensitivity to planets at large radii than do radial velocity searches. For
the VLBA and the expanded VLBA, the targets of radio astrometric surveys
are by necessity nearby, low-mass, active stars, which cannot be studied
efficiently through the radial velocity method, coronography, or optical
interferometry."

The addition of Arecibo in such surveys would increase the detection
sensitivity by a factor of four, making it possible to study objects with
one third the mass of Jupiter as companions of stars of similar types as in
RIPL. As Arecibo's primary beam is much smaller than that of other
telescopes, and the slew rate lower, the availability of a small antenna
for phase referencing would be highly beneficial for undertaking such
studies.

2.2 A Broad-impact VLBI Measurement of Trigonometric Parallax of Star
Clusters: In an impressive work using the VLBA at 8 GHz, Menten et al.
(2007, A&A, 474, 515) have determined the trigonometric parallax of several
stars in the Orion BN/KL region, allowing them to derive the most accurate
value to date (414± 7pc) for the distance to this region. This is about an
order of magnitude better than the previous value of 361+168-87 pc
determined by Hipparcos from the optical parallax measurement of a single
star in this complex. Luminosity-based distance estimates of star-forming
regions may be adversely affected by poorly known extinction. The new radio
technique is an important way to improve the estimation of distances, and
hence luminosities, with subsequent impact on star-formation theories.
Once again, the inclusion of Arecibo would permit the extension of these
studies to fainter, more distant, star-forming regions.

2.3 Pulsar Astrometry: High-precision astrometry of pulsars over multiple
epochs can provide their basic astrometric parameters: positions, proper
motions, and annual trigonometric parallaxes. Due to the weakness of most
pulsars, with duty cycles of typically <10%, the participation of Arecibo
with phase referencing is vital to the success of this exercise. In respect
of positional measurements, we note that VLBI estimations are tied to the
reference frame of the distant quasars, rather than the Solar-system frame
employed by pulsar timing positional estimates. This allows fundamental
reference frame ties between the Solar-system and extragalactic (ICRF)
frames via measurements of recycled pulsars, which are highly stable
rotators.

Proper motion estimates allow pulsars to be traced back to their birth
sites and, for very young pulsars, associations with progenitor supernova
remnants (SNRs) can be established, providing independent age estimates for
the SNRs. Combined with pulsar distance estimates, proper motion
measurements lead to estimates of space velocities, allowing a study of the
natal kicks imparted to pulsars at the time of their birth. When a parallax
measurement is possible, this yields a model-independent estimate for the
distance (and hence velocity) of the neutron star. Such measurements, (i)
calibrate models of the Galactic electron distribution, (ii) constrain SN
core collapse using the velocity estimates, and (c) provide photospheric
sizes for hot neutron stars with optically observed thermal surface
radiation, which in turn constrains the equation of state of matter at
extreme pressures and densities.

2.4 Detection Experiments: Present-day VLBI offers the highest sensitivity
radio astronomical observations yet achieved, with noise levels presently
approaching 1 ?Jy/beam for arrays using the world's most sensitive
telescopes. Hence, the 305-m Arecibo telescope is being increasingly used
in experiments to detect radio emission from very weak, very compact,
astronomical targets such as X-ray stars, distant supernovae and their
remnants, Gamma-Ray Bursts, and red-dwarf and other stars. For these
sensitivity levels to be reached for targets of very low intensity, it is
essential that phase-referencing be used.

2.5 VLBI Imaging of Molecular Gas in ULIRGs: Arecibo and the GBT are
currently searching for cm-wavelength lines of prebiotic and other
molecules in Ultra-Luminous InfraRed Galaxies (ULIRGs). The project has
been inspired by the recent Arecibo detection of the prebiotic molecule,
methanimine (CH2NH), in the protypical ULIRG/megamaser galaxy, Arp 220
(Salter et al. 2008, AJ, 136, 389). These galaxies are considered to be
"extreme mergers" and are heavily obscured at optical wavelengths.
Molecular lines from them often show wide velocity widths, caused by line
blending due to spatial and velocity overlaps. Detailed studies of maser
emission and molecular absorption lines from these objects require phase-
referenced VLBI observations, and the presence of Arecibo's sensitivity in
the VLBI array.

3. Astronomy Using the AT as an Independent Single Dish

3.1 Full-Stokes Galactic Plane Continuum Surveys: The small telescope,
together with existing Arecibo backends, will enable full-Stokes, continuum
surveys using the dual-channel receivers that will be used with the dish.
Full-Stokes continuum surveys of the wider Galactic plane at high
frequencies with the AT can provide unique databases in a number of ways.
Firstly, they can yield full spatial frequency, full-Stokes mapping at
previously unmapped wavelengths, with competitive resolution for such
extended features as the Galactic background emission, HII complexes, and
middle-aged and old SNRs. Comparison with existing lower-frequency surveys
would allow accurate estimation of spectral index distributions over these
features, providing the ability to perform accurate thermal-nonthermal
separation on angular scales between ~1o and 10 arcmin allowing the study
of energy injection to the ISM, energy losses for relativistic particles
associated with SNRs, and the mechanisms of vertical transport and
diffusion of energy from the disk of the Galaxy into the halo and
intergalactic space.

Linear polarization measurements are especially important. The appearance
of the polarized sky at ? > 21 cm is complex. Westerbork at 327 MHz (for
high Galactic latitudes) and the 1.4-GHz Canadian Galactic Plane Survey
have shown that there is little relationship between total intensity and
polarization structures; for the diffuse Galactic synchrotron emission, the
bulk of the Galactic Plane imaged at L-band reveals highly structured
polarization features with no Stokes I counterparts. The accepted
interpretation of this is that, although the Galactic synchrotron emission
is intrinsically quite smooth, differential Faraday rotation in the
intervening magneto-ionic medium, (the Faraday Screen), imposes fine
structure on the polarized emission, i.e. the low-frequency polarized sky
is dominated by propagation effects rather than intrinsic emission
structure. The signals produced via the Faraday Screen are rather weak, and
the limited surface brightness sensitivity of interferometers samples only
the strongest. Also, the derived rotation measures (RMs) are noisy due to
low signal-to-noise per channel, while missing zero-spacings in
interferometric observations lead to complications in interpretation.

The high brightness sensitivity of the AT, coupled with its few arcminute
beam size at high frequencies, promises major advances in the study of the
magneto-ionic medium. At these frequencies (™ 5 GHz), the effects of
Faraday rotation become tiny (?? ? ?-2) and the AT polarization position
angles will essentially be those intrinsic to the emission, providing both
directions of magnetic fields and a database against which lower frequency
polarization distributions can be definitively interpreted. In existing
studies of the Faraday Screen, the spectral signatures of the polarized
intensity have been examined to seek only a single RM value per image
pixel. Such a value corresponds to the RM of the dominant polarized
emission component along any given sight-line. However, the Faraday Screen
is spread out in depth along each line of sight, with regions of polarized
emission at different distances along the sight-line contributing to the
observed spectrum with their corresponding foreground Faraday rotation
signature. With an appropriate combination of observing frequency,
bandwidth and spectral resolution, it should be possible to perform Faraday
tomography, wherein the spectral polarized intensity modulations along a
given sight-line can be transformed to a set of polarized intensities as a
function of Faraday depth (i.e. RM), i.e. a polarized-intensity data cube
(quite like a spectral-line data cube) with two dimensions being the sky
coordinates and the third being RM. High-frequency images from the AT would
be invaluable in pursuing this endeavor.

Away from the Galactic plane, the high latitude regions contain several
well-known non-thermal emission structures, notably the North Polar Spur
(Loop 1), an object that contains rich small-scale structure, both on its
main arc and in internal ridging. Above b = 45 њ, low resolution
measurements of this nearby (~100 pc distant), old SNR show >70% linear
polarization at 1.4 GHz. Higher frequency, higher resolution AT images
will directly reveal the detailed magnetic field structure in this object.

We specifically mention the L-band Arecibo GALFA Continuum Transit Survey
(GALFACTS), which is being made by an international consortium led by Prof.
Russ Taylor (U. Calgary). This full-Stokes survey of the whole sky
observable with the 305-m telescope covers 1225 - 1525 MHz, with 8192
frequency channels. At L-band, the Faraday rotation effects on the
linearly polarized radiation are considerable, and a continuum survey at
much higher frequency, but similar resolution, (HPBW ~ 4 arcmin for
GALFACTS), would allow thermal-nonthermal separation, and aid Faraday
tomography when combined with GALFACTS. The same situation exists for a
large part of the Southern Galactic Plane L-band continuum survey being
made with the Parkes radio telescope (HPBW ~ 15 arcmin; Haverkorn et al.
2006, ApJS, 167, 230). AT surveys would also provide vital low-spatial
frequency data for future interferometric full-Stokes surveys.

Synergy with GLAST: ?-ray emission from our Galaxy is believed to be
produced by, a) brehmsstrahlung from the interaction of cosmic-ray
electrons and the interstellar gas, and b) the decay of neutral pions
produced in interactions between the gas and cosmic-ray protons and heavier
nuclei. The former is thought to dominate at <1 MeV, the latter at higher
energies. Similar distributions of ?-rays are found at low latitudes in
both energy ranges, suggesting that the cosmic-ray heavy particle-to-
electron ratio is constant over the Galaxy. If so, the ?-ray emissivity,
??, is proportional to the product of the cosmic-ray intensity and the
total (i.e. neutral and ionized atomic, plus molecular) gas density, ? ;
?? ? ?0 ?, where the cosmic-ray energy distribution is given by, N(E) dE =
N0 E-? dE. Now, for the synchrotron component of the Galactic radio
emission, the emissivity is, ? ? ?0 ?? (?+1)/2, where ?? is the
magnetic field strength perpendicular to the line of sight.

The Galactic distributions of the three quantities, ?0, ?, ?, are all of
great astrophysical interest. Arecibo will contribute significantly to a
knowledge of ? over the accessible sky, with the GALFA consortia providing,
a) the 2-dimensional distribution of HI, while b) the thermal-nonthermal
separation of the continuum emission mapped by the AT and GALFACTS surveys
will provide the 2-D distributions of the thermal emission from HII and
the non-thermal synchrotron emission. The 2-D distribution of the
molecular gas is already available from CO surveys of similar resolution.
Hence, combining Arecibo AT and ALFA results with other radio data and the
high-fidelity GLAST ?-ray background images will provide the information
needed to "unfold" the 2-D distributions and derive the Galactic
distributions of ?0, ?, & ?. This would represent a major contribution to
our understanding of the detailed distribution of the magnetic field and
cosmic rays in the Galactic disk.

3.2 Pulsar Timing: The majority of pulsars are too weak to be
satisfactorily timed with the proposed AT. However, timing observations of
the strong Crab and Vela pulsars would compliment dedicated monitoring of
the rotational state of these neutron stars undertaken at other
observatories. The traditional aim has been to detect glitches and timing
noise in these objects, which is scientifically important by itself (e.g.
Lyne et al. 1996 Nature, 381, 497 & Lyne et al. 1993, MNRAS, 265, 1003).
This has become even more relevant now with the launch of ?-ray satellites
such as AGILE and GLAST, which require continuous monitoring of these
pulsars in order to fold the many gamma-ray pulsar candidates with a
temporal and hour angle coverage that are impossible to achieve with the
305-m telescope, and that complement other telescopes: for instance, the AT
could time the Vela pulsar while it is below the horizon of the dedicated
telescope at Mt. Pleasant in Tasmania.

Another application of this antenna will be on the statistical properties
of the giant pulses of the Crab and other pulsars. These have fluxes of the
order of MJy, which are very easily detectable with the AT. This would have
several important ramifications. An international collaboration is using
several millisecond pulsars to detect very low-frequency gravitational
waves. One of the main issues of this project has to do with the
calibration of the time delays of all the telescopes and back-ends used for
this project. The giant pulses from the Crab detected by the AT can be used
as an international time reference for all telescopes in the Western
Hemisphere, giving a calibration of the instrumental delays of all the
telescopes and back-ends involved in this global high-precision timing
project.

The signals received by the AT will be piped back to the control room for
processing with existing Arecibo back-ends, which have well-known time
characteristics. This avoids duplication in the processing hardware or
software, although it will add to data storage and processing requirements.


3.3 Seven-Beam Interferometry with ALFA:

At L-band, the 7 ALFA beams of the 305-m dish would all lie near the peak
of the (voltage) primary beam of a small (say, 12-m class) antenna. As the
ALFA voltage beams have essentially double the sky coverage of their power
patterns, and overlap close to the half power points, one can foresee
"mosaicing" fair-sized pieces of sky from AT - ALFA interferometry. The
future addition of more elements at other sites around P.R. could
enhance this possibility, providing both uv-coverage, and phase and/or
amplitude self-calibration. The necessary cross correlator for the AT -
ALFA interferometry, i.e. 7 baselines, could likely be provided by an
existing backend, or be built collaboratively. Software correlators,
currently under development for eMERLIN and the VLBA, could also be used
at AO for this purpose. This wide-field interferometric mode of operation
would appear relevant to development work for the Square Kilometer Array
(SKA).

4. Geodetic use of the AT:

The International VLBI Service for Geodesy & Astrometry (IVS) recently laid
down a set of guide lines for the next-generation of geodetic VLBI
measurements of station positions and earth orientation parameters. These
are called VLBI 2010.The aim is to achieve 1 mm positional accuracy on
intercontinental baselines with a 24-hr turnaround for results. However,
there is insufficient funding for an international network at present, and
the geodetic community is interested in any telescope that could
participate. As is being planned for the AT, VLBI 2010 antennas need to
have minimal horizon obstruction, and be sited on good bed-rock.

The VLBI 2010 requirements specify a telescope design of 12-m diameter with
a system temperature of about TR = 45 K, SEFD < 2500 Jy. A frequency
coverage of 2 - 18 GHz will ultimately be needed. Dual-frequency operations
are a requirement to allow for ionospheric correction. An eventual data
recording rate of 8 - 16 Gbps is being set as a goal, and is consistent
with NAIC development plans for its VLBI backend. Recording at a rate of 4
Gbps was recently made with the 305-m telescope using a borrowed digital
backend, and dual state-of-the-art Mk5 recorders. Much of the data transfer
is envisaged as being over the internet (e-transfer), and Arecibo is
already the main player in American astronomical eVLBI.

Arecibo is attractive t0 VLBI 2010 as it is situated on the Caribbean
techtonic plate. This plate is very complicated, and accurate measurements
of velocities in a global frame are most valuable. Currently, the 25-m VLBA
antenna on St. Croix is available to the geodetic array, but only on 6
occasions per year. For each geodetic session, the geodetic community would
need 24 hr participation. The minimum involvement required is once per
month, though once per week would be welcomed. While the geodetic
community intend extending their frequency coverage to Ka band eventually,
this would not be necessary for some while. VLBI 2010 would require 2-bit
sampling, and initial recording rates of (say) 2 Gbps.

To participate in geodetic VLBI, a cable-delay system (i.e. A "round-trip"
phase measurement) will be needed as system delays need to be tracked at
the mm-level. Diurnal effects should be monitored, and the receiver has to
be as stable as possible. A dual-frequency GPS receivers should exist
close to the antenna.

Geodetic operations are the responsibility of the individual stations.
Analysis centers exist around the World, with some being based at
universities. It is likely that UPR could become involved. We note that the
analysis of VLBI data is simple compared to that for GPS data. Further,
while GPS is good for measuring regional motions (over ~1000 km), VLBI can
provide the tie to more distant points with better accuracy than GPS.

5. An Additional Benefit - the Geographical Location of the 305-m Telescope

The geographical position of the Arecibo 305-m radio telescope has always
been difficult to determine with the accuracy required for VLBI. This is in
large part due to the very limited range of zenith angle (<19.7њ) that the
telescope can access. The currently-used position was derived by Dr. David
Graham of the MPIfR from the data acquired for an actual VLBI experiment in
2001. This is believed to only be good at the "meter level". In 2002 we
explored with the teams at NRAO and GSFC the possibilities for improving
this to centimetric accuracy. The bottom line here was that the best
accuracy for this measurement could be obtained if a smaller telescope with
full sky coverage were to exist at Arecibo for which a full VLBI "ties"
experiment could establish a similar accuracy geographical location. Local
surveying techniques would then be used to establish a high accuracy
position for the 305-m telescope. This poor knowledge of the 305-m
telescope position raised its head again recently following the pioneering
Ultra-wideband VLBI (UVLBI) experiment performed using Arecibo, the GBT and
the VLBA by scientists from the MIT Haystack Observatory. The results show
that on all Ar baselines the residual delays are time variable. This is
almost certainly due to uncertainties at the tens of cm level in the
telescope's position. Eliminating such uncertainties is crucial for the
quality of images obtained from phase-referenced observations, and the
arrival of an auxilliary telescope would facilitate obtaining the best
possible position for the world's largest telescope.

6. Educational efforts involving the AT

NAIC is entering into a collaboration with the UPR-Humacao for integrating
the AT for their undergraduate physics/electronics teaching and research
plan. To this end, a funding proposal was submitted to the NSF-PAARE
programe in August 2008.



7. Technical Requirement for the small Auxiliary Telescope

In this section, we describe what is needed to achieve the above
objectives.

7.1 How Big an Antenna is needed for Phase Referencing?: The typical
calibrator used for phase-referenced VLBI is > 100 mJy/beam, a value that
includes the majority of the sources from the VLBA Calibrator List. Hence,
an antenna size is needed that will give an acceptable signal-to-noise
ratio for sources of this intensity. This implies that for successful phase
referencing the rms (1-?) noise on a baseline would have to be > 20
mJy/beam in the typical integration time on a phase-referencing calibrator.
Now, for a "worst-case scenario" we consider a baseline between an Arecibo
auxiliary antenna and a 25-m VLBA antenna. For a single polarization of 16-
MHz bandwidth, observing a source for a typical dwell time of 120 sec, the
size of dish, D, of aperture efficiency, ?, and system temperature, T,
needed to give an rms noise of ? mJy/beam is given by D2 = 1.435*(T/?).
For T = 30 K, ? = 0.7, (parameters expected for a Patriot dish, equipped
with a cryogenic receiver), D = 7.8 m. For values of T = 50 K, ? = 0.6,
(parameters appropriate for the DSS33 antenna described below, mounted with
either a good room temperature receiver, or a possible wide-band feed), D =
10.9 m. Hence, a 12-m class antenna would do the job adequately, although
a larger antenna would, of course, give improved signal-to-noise ratio for
VLBI by a factor of (D/12.).

7.2 Required Frequency Coverage: For VLBI phase referencing requirements,
it would be essential to cover the standard VLBA C- and X-band continuum
frequency bands near 5.0 and 8.4 GHz. Coverage of the standard VLBA L-band
frequencies at 1.65 and 1.42 GHz would also be welcome. For participation
in the VLBI 2010 geodetic program, simultaneous coverage of S and X bands
(2.4 and 8.4 GHz) is presently required. It would seem optimum to begin
operations with either an S/X band receiver or a locally produced C-band
receiver, using this to commission the dish, develop and implement the
novel approach to VLBI phase referencing that we propose to employ using
the AT and the 305-m telescope, and to allow the use of the telescope for
single-dish research and educational purposes. However, the number of
frequency bands required would best be accommodated in the longer term
using a wide-band feed system (say 1 - 10 GHz) as is presently being
developed by German Cortes of NAIC, Ithaca. We note that the system
temperature of such a wideband feed is expected to be somewhat higher than
that for a cooled narrow-band receiver. However, provided the system
temperature does not much exceed 50 K, even a 12-m diameter dish should
meet our requirements for VLBI phase referencing.

7.3 Potentially Acceptable Auxiliary Antennas: In searching for an
economical solution to obtaining an "auxiliary" antenna for NAIC to use for
phase-referenced VLBI (and other roles), we will consider here the
following possibilities;

a) We learned recently that NASA/Canberra Deep Space Communication Complex
(CDSCC) had moth-balled their 11.3-m antenna (DSS-33), earlier used as a
HALCA/VSOP earth station, and were looking for somebody who would be
interested in taking over the dish. We have expressed our interest to NASA.


b) An alternative solution would be the acquisition from Patriot Antenna
Systems of either their 12.1-m antenna that has been adopted as standard
for the VLB2010 project (see Section 4), or even one of their larger
designs, which range up to 18.3-m diameter.

We summarize the expected radiometric performance of these possible
solutions as follows;

Antenna Aper. Eff. K/Jy SEFD(Jy) HPBW5GHz ?conf
(5 Ghz)

NASA DSS-33 64% 0.023 43.5*Tsys 23 arcmin
29 mJy
Patriot 12.1-m 69% 0.029 34.5*Tsys 21.5 arcmin
25 mJy
Patriot 18.3-m 71% 0.067 14.8*Tsys 14 arcmin
11 mJy

(The Aperture Efficiencies are calculated from available figures for
forward gain. The SEFD values are given for a system temperature of Tsys,
HPBW5GHz is the Half Power Beamwidth at 5 GHz, and ?conf the rms confusion
at that frequency.)

All three antennas would work well radiometrically to a frequency of at
least 15 GHz, (possibly higher; see below), meaning that they will operate
efficiently across the complete frequency range of the 305-m telescope
(i.e.