Summary

During the nights of September 15/16 and 16/17, 2002, preliminary tests
were successfully carried out during which the light beams from all four
VLT 8.2-m Unit Telescopes (UTs) at the ESO Paranal Observatory were
successively combined, two by two, to produce interferometric fringes.
This marks a next important step towards the full implementation of the
VLT Interferometer (VLTI) that will ultimately provide European
astronomers with unequalled opportunities for exciting front-line research
projects.

It is no simple matter to ensure that the quartet of ANTU, KUEYEN, MELIPAL
and YEPUN, each a massive giant with a suite of computer-controlled active
mirrors, can work together by sending beams of light towards a common
focal point via a complex system of compensating optics. Yet, in the span
of only two nights, the four VLT telescopes were successfully “paired” to
do exactly this, yielding a first tantalizing glimpse of the future
possibilities with this new science machine.

While there is still a long way ahead to the routine production of
extremely sharp, interferometric images, the present test observations
have allowed to demonstrate directly the 2D-resolution capacity of the
VLTI by means of multiple measurements of a distant star.

Much valuable experience was gained during those two nights and the ESO
engineers and scientists are optimistic that the extensive test
observations with the numerous components of the VLTI will continue to
progress rapidly. Five intense, technical test periods are scheduled
during the next six months; some of these with the Mid-Infrared
interferometric instrument for the VLTI (MIDI) which will soon be
installed at Paranal.

Later in 2003, the first of the four moveable VLTI 1.8-m Auxiliary
Telescopes (ATs) will be put in place on the top of the mountain; together
they will permit regular interferometric observations, also without having
to use the large UTs.

Less than one year after the first combination of two 8.2-m VLT telescopes –
described in detail in ESO Press Release 23/01 – successful tests have now
been carried out, during which all of the four telescopes were combined
pairwise in rapid succession.

Of the six combinations possible (ANTU-KUEYEN, ANTU-MELIPAL, ANTU-YEPUN,
KUEYEN-MELIPAL, KUEYEN-YEPUN and MELIPAL-YEPUN), only the last one could not
be used, because of the current geometrical configuration of the three delay
lines installed so far.

The combination of the light beams from two (or more) VLT Unit Telescopes is
a daunting task. It involves pointing them simultaneously towards the same
celestial object, ensuring optimal optical adjustment of the
computer-controlled telescope mirrors (including the shape of the 8.2-m
primary mirror by “active optics”), performing extremely smooth and stable
tracking of the object as the Earth turns, guiding the light beams via
additional (“coude”) mirrors into the “delay lines” installed in the
Interferometric Tunnel below the telescope platform, keeping the total path
lengths equal to within a fraction of a micron during hours at a time and
finally, to register the interferometric fringes at the focal point of the
VINCI instrument [1], where the light beams encounter each other.

Next year, the first adaptive optics systems for the VLTI will be inserted
below the telescopes. By drastically reducing the smearing effects of the
turbulent atmosphere through which the light has to pass before it enters
the telescopes, this will further “stabilize” the imaging and increase the
sensitivity of the VLTI by a factor of almost 100.

At this moment, three delay lines have been installed, but for the present
first test, the VLTI engineers and astronomers used the telescopes in pairs,
in order to set-up the various equipment configurations properly. In this
way, they could also start “teaching” the computer control software to
handle this very demanding process as efficiently and user-friendly as
possible in the future. With the arrival of the science instrument AMBER in
mid-2003, up to three beams can be combined simultaneously.

It turned out that the various predictions of mirror positions and angles
were quite accurate and only a moderate amount of time was needed to “obtain
fringes” in all different configurations. Measurements were then made on a
number of stars, among them the brightest star in the southern constellation
Eridanus (The River), known as Alpha Eridani or Achernar, that was observed
several times with the different telescope pairings. This star is a hot
dwarf (spectral type “B5 IV”) that is located at a distance of about 145
light-years. It has also been extensively observed during earlier VLTI
tests. It is a very suitable object for the present resolution tests as its
angular diameter is only about 0.002 arcsec and it therefore remains
unresolved at the near-infrared wavelength of the K-band used (2.2 micronm).

In fact, the combination of these data (including also some that were
obtained in October 2001) now makes it possible to reconstruct the first
interferometric “point-spread function (PSF)” of a star obtained with the
VLTI, cf. PR Photo 22b/02. This is like an “interferometric image”, except
that the disk of this particular star remains unresolved.

The angular resolution is inversely proportional to the aperture of a
telescope for single telescope observation, and to the length of the
“baseline” between two telescopes for the interferometric observation.
However, observing interferometrically with two telescopes will improve the
resolution only in the direction parallel to this baseline, while the
resolution in the perpendicular direction will remain that of a single
telescope. But then the use of other telescope pairs with different baseline
orientations “adds” resolution in other directions.

The reconstructed PSF of Achernar shown in PR Photo 22b/02 is obviously
still very incomplete, due to the technical nature of the present tests and
the limited time that was spent observing the star in each configuration.
However, it already presents a powerful illustration of the extreme imaging
sharpness that will be achieved with the VLTI.

Note

[1]: The VINCI instrument was built under ESO contract at the Observatoire
de Paris (France) and the camera in this instrument was delivered by the
Max-Planck-Institute for Extraterrestrial Physics (Garching, Germany). The
IR detector and the IRACE detector electronics were supplied by ESO.