eso0134 — Organisation Release

Giant Eyes for the VLT Interferometer

First Scientific Results with Combined Light Beams from Two 8.2-m Unit Telescopes

5 November 2001

It started as a preparatory technical experiment and it soon developed into a spectacular success. Those astronomers and engineers who were present in the control room that night now think of it as the scientific dawn of the Very Large Telescope Interferometer (VLTI). On October 29, 2001, ANTU and MELIPAL, two of the four VLT 8.2-metre Unit Telescopes at the ESO Paranal Observatory, were linked for the first time. Light from the southern star Achernar (Alpha Eridani) was captured by the two telescopes and sent to a common focus in the observatory's Interferometric Laboratory.

Following careful adjustments of the optical paths, interferometric fringes were soon recorded there, proving that the beams from the two telescopes had been successfully combined "in phase". From an analysis of the observed pattern (the "fringe contrast"), the angular diameter of Achernar was determined to be 1.9 milli-arcsec. At the star's distance (145 light-years), this corresponds to a size of 13 million km. The observation is equivalent to measuring the size of a 4-metre long car on the surface of the Moon.

This result marks the exciting starting point for operations with the Very Large Telescope Interferometer (VLTI) and it was immediately followed up by other scientific observations. Among these were the first measurements of the diameters of three red dwarf stars ("Kapteyn's star" - HD 33793, HD 217987 and HD 36395), a precise determination of the variable diameters of the pulsating Cepheid stars Beta Doradus and Zeta Geminorum (of great importance for the calibration of the universal distance scale), as well as a first interferometric measurement of the core of Eta Carinae, an intriguing, massive southern object that may possibly become the next supernova in our galaxy.

This milestone is another important step towards the ultimate goal of the VLT project - to combine all four 8.2-metre telescopes into the most powerful optical/infrared telescope system on Earth. When ready, it will be able to reveal at least 15 times finer details in astronomical objects than what is possible with any existing, single ground-based telescope.

First VLTI observations with two 8.2-metre telescopes

At 1 o'clock in the morning of October 30, 2001, ESO astronomers and engineers working in the VLTI Control Room successfully combined the light from ANTU and MELIPAL, two of the four 8.2-metre VLT Unit Telescopes at the Paranal Observatory. The same night, a series of high-resolution test observations with the VINCI instrument [1] at the focus of the VLT Interferometer (VLTI) proved that this complex system was functioning extremely well, and within the technical specifications.

Following about seven months after the moment of "VLTI first light" during which the light beams from two small test telescopes were combined - as described in detail in eso0106 - this accomplishment above all serves as a demonstration of the possibilities and potential of interferometric observations with the four giant VLT telescopes.

The two large telescopes used for the present test are separated by 102 metres. In order to properly combine the starlight received by them, a train of 25 mirrors is needed. All of them must be adjusted with a precision of one thousandth of a millimetre or better.

As can be seen on ESO Press Photo eso0134, the light from the observed star is first directed towards the Nasmyth focus by three mirrors in the telescope tube. From here, it continues towards the intermediate Coudé focus below the telescope and then onwards through a subterranean light duct to the VLTI Delay Lines that are installed in the Interferometric Tunnel. At the end of this long chain of mirrors and after traveling a distance of approximately 200 metres, the light finally reaches the VINCI instrument in which the two beams interact coherently (in phase) to produce "interferometric fringes".

The tests have shown that the starlight arrives at the VINCI instrument with a pointing accuracy of about 1 arcsecond and, even more important, with a long-term tracking stability of the order of 0.2 arcseconds per hour.

In fact, the image quality measured at the focus of VINCI is essentially identical to that of the individual telescopes at the Nasmyth (and Cassegrain) foci. Stellar images as sharp as 0.4 arcsec (note that this is the size of the "seeing disk" FWHM, not yet a real image of the stellar surface; the VLTI will start producing two-dimensional images of stars and other objects at a later stage) have been obtained at the interferometric focus, c.f. ESO Press Photo eso0134. The installation of an Adaptive Optics system (see below) will later reduce the image size to the theoretical limit of 0.057 arcsec (for observations with an 8.2-m telescope in the infrared K-band at wavelength 2.2 µm (or 0.032 arcsec in the J-band at 1.2 µm).

First scientific results already during the test observations

The technical demonstration being so successful, the ESO astronomers and engineers involved in the development of the VLTI immediately decided to go one step further. And indeed, the interferometric fringes recorded with the light beams from two 8.2-metre VLT telescopes during these initial technical tests have already led to some very valuable scientific results.

The first star to be observed - the brightest star in the southern constellation Eridanus (The River) and known as Alpha Eridani or Achernar - is quite different from our Sun. It is estimated to be several times more massive and, with a surface temperature of about 20000 degrees, it is about three times hotter than our local star.

The distance to Achernar has been measured by the ESA HIPPARCOS satellite as about 145 light-years, and from its apparent brightness, it is found to be almost 1000 times more luminous than the Sun. Consequently, it depletes its energy resources much faster and has a much shorter life expectancy (about 100 million years) than the Sun (about 10,000 million years).

The new measurement with the VLTI found the angular diameter of Achernar to be 0.00192 ± 0.00005 arcsec . This is equivalent to the angle subtended by a 1 Euro coin (diameter 23.25 mm) as seen from a distance of 2500 km, or by a car (4 metres long) on the surface of the Moon. At the indicated distance, this angle also shows that the real size of Achernar is about 13 million kilometres, and that it is therefore nearly ten times larger than our Sun.

Following that first observation, and in spite of the many technical tests scheduled at this moment of the VLTI commissioning work, the astronomers were able to carry out several other scientific observations. During this exciting first period of operation, among others, measurements were made of three red dwarf stars, three stars surrounded by disks, one red giant star, two Cepheid stars and one luminous blue variable star. Preliminary results from some of these observations are described in the Appendix.

Angular measurements with the VLTI like the present ones will soon become routine and will allow astronomers to measure accurately the physical characteristics of many different types of stars. For instance, the precise measurement of the angular diameter of Achernar will make it possible to deduce directly and accurately its surface temperature, an important information for our understanding of the formation and evolution of such hot and massive stars.

From 40-centimetre to 8.2-metre

The present event follows after half a year of much hard work by ESO astronomers and engineers.

Earlier this year, the VLTI achieved "first fringes" by combining two small 40-cm siderostat telescopes. Since then, ESO astronomers and engineers have upgraded the VLTI and are preparing it for regular observations that will start next year. The present results obtained with the combination of two giant telescopes constitute one important milestone along this road.

Between March and October 2001, about 1000 individual measurements were carried out on celestial objects with the light beams from the small test telescopes. This process is on-going, as part of the commissioning of the VLTI, and is aimed at a detailed technical characterization of the interferometer and thorough knowledge of its performance. Such observations mainly serve to obtain technical data.

Nevertheless, some of them also provide interesting scientific results . For example, during the week just prior to the first fringes now achieved with two large telescopes, nearly 150 measurements were obtained over 4 nights. Among them, five Mira stars (a type of large and cool, pulsating stars) and two close binary stellar systems were observed - some of them had never before been studied interferometrically. Moreover, a large number of objects were observed for calibration. These data are now being evaluated, and will help astronomers to refine their understanding of the capabilities of the VLTI - they will soon become available to the astronomical community via the VLT archive.

In the same period, substantial additions were made to the system, e.g., a third Delay Line was installed in the Interferometric Tunnel. This allows the use of the telescopes on the east side of the beam combination laboratory (including MELIPAL) and also to combine the light beams from up to three telescopes at a later moment. The additional mirrors needed in order to permit the combination of the light from the two 8.2-metre telescopes were installed.

The extensive software that controls the telescopes and the instruments has undergone several revisions to accommodate the increased needs required by the more complex system of Unit Telescopes, delay lines and test instruments. At the same time, the overall reliability of the facility has been constantly improved.

The path that the light travels from the two 8.2-metre telescopes to the VINCI instrument must be kept constant to within a fraction of a micron, or better than one thousandth of a millimetre! Although it is therefore extremely sensitive to even very small disturbances, the VLT Interferometer has proven to be remarkably reliable and robust. For instance, an earthquake of magnitude 4+ on the Richter scale happened in August 2001 in the middle of a series of interferometric measurements. However, thanks to the many safeguards and compensatory measures built into the system, the VLTI continued to function all through the tremor. The observations were barely affected by the ground vibrations.

It should also be noted that, unlike the 40-centimetre siderostat telescopes, the 8.2-metre telescopes are so large that the images they produce are significantly affected by atmospheric turbulence. In order to overcome this problem, ESO is now developing a system of "Adaptive Optics" correctors (MACAO) which will "remove" the distortions introduced by the atmospheres by means of small, rapidly reacting computer-controlled deformable mirrors. From 2003, this system will increase the sensitivity of the VLTI by a factor of about 100 (5 magnitudes) compared to the present observations without adaptive optics.

VLT Instrumentation

The next steps in the VLTI project will be the integration of a new instrument working at a wavelength of 10 µm (the Mid-Infrared interferometric instrument for the VLTI (MIDI)) in the middle of 2002, the addition of a fringe tracker (FINITO) and then of a 3-way, 3-photometric bands instrument (the near-infrared/red VLTI focal instrument (AMBER)) at the beginning of 2003.

Following closely will be the addition of three 1.8-metre movable telescopes dedicated to interferometry, and of the Adaptive Optics system. With all these components in place, the VLTI will represent the most powerful interferometer available in the southern hemisphere, and will enable scientific investigations on a wide range of topics ranging from the direct detection of planets around other stars, to the formation and early evolution of stars, to the study of extragalactic objects.

A dedication to Ariela Rijo

On behalf of the staff, the Director of the Paranal Observatory adds this message:

"The Paranal Observatory, while very pleased at the present success of the first fringes from two of the 8.2-m telescopes, at the same time is greatly saddened by the loss of our colleague Ariela Rijo who passed away on October 31".

"She was a wonderful person and an excellent colleague who contributed greatly to the implementation of the VLTI on Paranal. The Paranal Observatory dedicates this result to her memory".


[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 MPI for Extraterrestrial Physics (Garching, Germany). The detector and the detector electronics was supplied by ESO.

More information

Scientific Appendix: First VLTI stellar measurements with two UTs

This appendix presents some technical details of the measurements, obtained with the VLTI and two UTs during the first three test nights. While it must be emphasized that the stated results are still provisional, they clearly indicate the excellent performance of the VLTI already at this early stage and, not least, the great potential for important fundamental observations with this facility. Note in particular, that the quoted errors reflect the statistical uncertainty in the data only and that additional calibration errors must later be taken into account.

The observational data were taken on a variety of astronomical objects, including three red dwarfs, three stars surrounded by disks, one red giant, two Cepheids and one luminous blue variable. All of these measurements were calibrated by observing a reference star of known angular size. Each data set required about ten minutes of continuous observations.

Fringes were found on all pointed objects within a few minutes of time and kept for up to several hours. All data were deemed to be of high quality and will be analyzed in detail within the next weeks. A preliminary data reduction was possible for part of these objects and it gave the results listed below (all quoted values are uniform disk diameters):

  • For the blue dwarf Alpha Eridani, on which first fringes were found, 11 data sets were taken within three nights and an angular diameter of 1.92 ± 0.05 milli-arcsec (mas) could be estimated, which is precisely in line with previous measurements.
  • The nearby red dwarf HD 217987 was measured to have a diameter of 0.92 ± 0.05 mas, resulting from two data sets. This is the first measurement of the angular diameter of a star as small as a type M0 dwarf, and one of the very few available for cool main sequence stars in general.
  • The giant star HD 36167 was found from four data sets to have a diameter of 3.32 ± 0.02 mas. This measurement constitutes a significant refinement of the earlier, indirect estimate of 3.55 ± 0.06 mas (Cohen M. et al. 1999, Astronomical Journal 117, 1864).
  • For the three stars which are known to be surrounded by a disk, the following results were obtained: Epsilon Eridani 2.20 ± 0.02 mas (8 data sets in two nights); Fomalhaut (Alpha Piscis Austrini) 2.31 ± 0.02 mas (4 data sets); Beta Pictoris unresolved (4 data sets). Further analysis is expected to put a significant lower limit on the visibility for the latter star.
  • The two Cepheids Zeta Geminorum and Beta Doradus showed diameters of 1.78 ± 0.02 mas (7 data sets) and 2.00 ± 0.04 mas (6 data sets), respectively. The diameter of Zeta Geminorum has been measured before by three different interferometers. Its diameter is expected to vary between about 1.5 mas and 1.8 mas within ten days. On the date the VLTI data was taken, its phase was close to the foreseen maximum diameter. Beta Doradus has never been measured before.
  • The red giant Psi Phoenicis was measured on two nights (16 data sets) with the UTs for three different positions in the sky, hence with three different projected baselines. Some weeks earlier it had been measured on four nights with the 40-cm test siderostats (8 data sets) on a shorter 16-m baseline. The star was well resolved already in the previous measurements, but the addition of the data recently obtained with the UTs is of fundamental importance because with their longer baseline and larger light-gathering power, it now becomes possible to obtain visibility measurements beyond the first null, cf. ESO Press Photo eso0134 . Such measurements in the future will enable astronomers to measure fine details such as limb-darkening and deviations from spherical symmetry. The preliminary diameter value for this star is 8.21 ± 0.02 mas.
  • The enigmatic object Eta Carinae is a luminous blue variable, a supermassive star, which underwent a massive outburst in the 1840's. This outburst was responsible for the creation of the surrounding Homunculus Nebula . The central object is not well understood, but is likely to have a complex structure and therefore the first interferometric measurement with the VLTI is of great importance. Fringes with a low contrast (amplitude of about 20%) were detected, indicating that the central object is resolved on a scale of a few milliarcseconds. More observations will be obtained to further investigate this peculiar object.


Andreas Glindemann
Garching, Germany
Tel: +49-89-3200-6590

Francesco Paresce
Garching, Germany
Tel: +4989-3200-6297

Connect with ESO on social media

About the Release

Release No.:eso0134
Legacy ID:PR 23/01
Name:Achernar, Instrumentation, Interferometry, Psi Phoenicis, Very Large Telescope Interferometer
Type:Unspecified : Technology : Observatory : Telescope
Facility:Very Large Telescope Interferometer


The VLT Interferometer with ANTU and MELIPAL
The VLT Interferometer with ANTU and MELIPAL
"Joint" Stellar light-spot
Interferometric fringes from the star Achernar
Interferometric fringes from the star Achernar
Time sequence of fringes from Achernar
Time sequence of fringes from Achernar
"Visibility Curve" of the star Psi Phoenicis