«Dissertation der Fakult¨ t f¨ r Physik au der Ludwig-Maximilians-Universit¨ t M¨ nchen au vorgelegt von Rainer Sch¨ del o M¨ nchen, den ...»
High Resolution Near-Infrared
Imaging Observations of the Galactic
Dissertation der Fakult¨ t f¨ r Physik
Ludwig-Maximilians-Universit¨ t M¨ nchen
vorgelegt von Rainer Sch¨ del
M¨ nchen, den 9.Februar 2004
Tag der m¨
ufung: 3. Juni 2004
1.Gutachter: Prof. Dr. Reinhard Genzel
2.Gutachter: Prof. Dr. Ralf Bender
Figure on cover page: Orbits of stars around Sagittarius A*. Lucy-Richardson deconvolved and beam-restored high-resolution (FWHM∼ 60 milliarcseconds) near-infrared (2.2µm) image of the central 1 × 1 of the nuclear stellar cluster around the supermassive black hole Sgr A*. The image was obtained with CONICA/NAOS at the ESO VLT in June
2003. On this image Sgr A* can be seen in its ﬂaring state as a point source at the origin of the coordinate system. The Keplerian orbits of six stars, as they were determined in this thesis, are overplotted onto the image. Arrows indicate the locations of the respective stars and their direction of motion.
Zusammenfassung Ziel der vorliegenden Doktorarbeit war es, neue Erkenntnisse ¨ uber die Struktur, Zusammensetzung und Dynamik des zentralen Sternhaufens unserer Milchstraße zu gewinnen. Im Mittelpunkt unserer Analysen stand dabei vor allem die Natur der Konzentration einiger Millionen Sonnenmassen dunkler Materie im Zentrum dieses Haufens, bei welcher es sich vermutlich um ein supermassives Schwarzes Loch handelt. Schon seit Jahrzehnten wurde vermutet, dass die kompakte, nicht-thermische Radioquelle Sagittarius A* (Sgr A*), welche 1974 entdeckt wurde, mit einem solchen Objekt assoziiert ist. In großen Teilen basiert diese Arbeit auf Beobachtungen des galaktischen Zentrums mit der neuartigen Nahinfrarotkamera CONICA und dem dazugeh¨ origen System f¨ adaptive Optik, NAOS, am Very Large Telescope der ur Europ¨ aischen S¨udsternwarte. Dieses kombinierte System wurde Ende 2001/Anfang 2002 in Betrieb genommen und bietet ideale Voraussetzungen f¨ tiefe, hochaufgel¨ Nahinfrarotur oste Beobachtungen des galaktischen Zentrums.
Ein grundliegendes Problem, welches es zu l¨ osen galt, war die Astrometrie der Aufnahmen des Sternfeldes im galaktischen Zentrum. Ein akkurates astrometrisches System ist eine essentielle Voraussetzung daf¨ Sgr A* auf Infrarotbildern zu identiﬁzieren und die relativen ur, Positionen und Bewegungen der Sterne in seiner Umgebung zu messen. Mit Hilfe von SiO Maser Sternen, deren Position durch Radiointerferometrie zu 1 Millibogensekunde
Chapter 1 Introduction The discovery of quasars in the second half of the 20th century and subsequent detailed research on these extraordinary objects soon led to the insight that these sources are the objects with the highest average energy output in the universe. Moreover, it was found that quasars produce their intense radiation on very small scales, i.e. are extremely compact sources. It was found that all observations can be explained best by the assumption that quasars are supermassive black holes of up to several billion solar masses located at the centre of galaxies that are accreting matter at a rate comparable to 1M yr−1 from their surroundings. While most quasars can be found at redshifts of roughly 2-3, they are much rarer in the local, i.e.
If supermassive black holes can be observed in the early days of the universe, they must still be present in the present-day universe. In fact, active galactic nuclei (AGN), such as in Seyfert galaxies, observed in the local universe, are considered scaled-down versions of quasars. If one assumes that the reason for the scarcity of quasars in the local universe is that they ran out of fuel, it should be possible to observe these enormous concentrations of dark matter at galactic nuclei via their inﬂuence on the dynamics of the surrounding stellar clusters or galactic bulges. Indeed, high-resolution spectroscopic observations with the Hubble Space Telescope and ground based telescopes as well as radio interferometric observations of maser disks have provided considerable evidence for the existence of supermassive black holes at the centre of quiet galactic nuclei, with 30 − 40 known candidates up to date (Kormendy, 2001, 2003). In the cases of our own galaxy, of M31, and of NGC 4258 the measurements exclude clusters of dark stellar remnants as explanations for the observed dark mass concentrations (Kormendy, 2003).
The centre of the Milky Way is about 100 times closer to Earth than the next major galaxy, M31, and 1000 times closer than the next AGN. Therefore, we can observe the phenomena at the centre of our home galaxy at a level of detail that can never be reached in the case of external galaxies. This makes the Milky Way an ideal candidate for testing the standard paradigm that supermassive black holes reside at the centre of most, if not all, galaxies.
The solar system is located in the galactic plane at a distance of about 8 kpc from the Galactic Centre (GC) (Reid, 1993). Dust clouds in the galactic disk therefore cause an extinction of about 30 magnitudes toward the GC at visual wavelengths. Hence, the GC can only be observed in the radio, infrared and X-ray regimes. Lynden-Bell & Rees (1971) ﬁrst suggested that the Milky Way might contain a supermassive black hole in its centre. When Balick & Brown (1974) detected the compact non-thermal radio source Sagittarius A* (Sgr A*, naming by Brown, 1982) in the GC, this object soon became the primary suspect for being the manifestation of such a black hole. Subsequent radio interferometric observations showed that the source is very compact, i.e. less than 1 AU in diameter (Rogers et al., 1994; Krichbaum et al., 1998; Lo et al., 1998; Doeleman et al., 2001). Backer & Sramek (1999) and Reid et al. (1999) inferred from multi-epoch observations that Sgr A* has a velocity projected on the
sky of 20 km/s. Compared to the velocity dispersion in the nuclear stellar cluster of several hundred km/s this means that Sgr A* must have a mass of at least several thousand solar masses.
The standard technique in the quest for testing the black hole hypothesis is measuring the gravitational potential at ever shorter distances to the black hole candidate Sgr A*. The amount of dark mass that is enclosed in a given volume can be determined from the dynamics of gas and stars in the galactic nucleus. First dynamical observations of the enclosed mass in the central parsec suggested the presence of a point-like object of ∼ 3 × 10 6 M at the heart of the Milky Way (see Genzel & Townes, 1987). However, these ﬁrst estimates still depended sensitively on the motion of ionised gas streamers (Serabyn & Lacy, 1985; Serabyn, 1988), which can be subject to forces other than gravity, such as magnetic ﬁelds or stellar winds.
Stars, however, would be the ideal test particles for measuring strong, large-scale gravitational ﬁelds.
For this undertaking, it was necessary to be able to measure the proper motion velocities of the stars in the GC cluster, i.e. their velocities as seen projected on the plane of the sky.
However, this required the capability of measuring their positions with a precision of a few milli-arc-seconds. This became possible with the availability of large near-infrared (NIR) detector arrays in combination with high resolution NIR imaging techniques at the beginning of the 1990’s, which opened the way for imaging observations of the nuclear cluster of the Milky Way at the diffraction limit of 4–10 m class telescopes.
High-resolution imaging techniques at NIR wavelengths involve basically two methods, speckle imaging and adaptive optics. Atmospheric turbulence causes a deterioration of astronomic images so that even under excellent conditions, the achievable resolution of long exposures is not better than about 0.5. The speckle imaging technique takes advantage of the fact that information at the diffraction limit of large telescopes is conserved if one chooses a short enough integration time, of the order 0.1 s. Series of short speckle exposures can be used for a subsequent restoration of a diffraction limited image with the help of a computer.
As for adaptive optics, brieﬂy explained, this technique allows measuring and correcting the major atmospheric image distortions in real time. A so-called wavefront sensor (WFS) in combination with a real-time computer serves to analyse the light of a bright star, that is taken as a reference point source, and sends a signal to a de-formable mirror that corrects the wavefronts of the incoming light before it is registered by the detector. The biggest advantage of the AO technique over speckle techniques is that it does not require short integration times of just fractions of a second. Hence, integration times of several seconds up to hours are possible.
This signiﬁcantly enhances the sensitivity of the observations because they are not limited any longer by the read-out noise of the detector.
The NIR speckle camera SHARP (Hofmann et al., 1995), built at the Max-Planck-Institut f¨ extraterrestrische Physik, was the ﬁrst such instrument that was dedicated to GC observaur tions. It was used at the ESO 3.5 m NTT telescope in La Silla, Chile, and provided the ﬁrst diffraction limited images of the central parsec of the GC (Eckart et al., 1992). This technology opened the way for measuring stellar proper motions near Sgr A*, the suspected supermassive black hole. Several epochs of observations accumulated indeed strong evidence that the gravitational potential in the central 0.5 pc of the Milky Way was dominated by a point mass of 2 − 3 million solar masses (Eckart & Genzel, 1996; Genzel et al., 1997, 2000). In 1995, a similar experiment was started by a group at the University of California Los Angeles (UCLA) with the 10 m Keck telescope, using also speckle imaging and later also adaptive optics (AO) techniques. They conﬁrmed the results that were found by the MPE group and also detected the ﬁrst evidence for accelerations of stars near the suspected supermassive black hole (Ghez et al., 1998, 2000).
At the end of 2001, the ESO VLT unit telescope 4 (Yepun) was equipped with the NIR camera CONICA and the AO system NAOS (Lenzen et al., 1998; Rousset et al., 1998). The combined instruments (“NACO”) serve to obtain diffraction limited NIR images on the 8m class VLT telescope. While conventional AO systems rely on guiding stars that are bright at Introduction 11 visible wavelengths for the wavefront sensing, NAOS has the unique feature of an infrared wavefront sensor (IWFS). This means that it can alternatively lock the AO on sources that are bright in the near-infrared regime if no visibly bright star is near the observed target. In the case of the GC, the nearest visible guiding star is more than 30 away from the target and rather faint (∼ 14 mag). This allows only moderate image corrections under good atmospheric conditions. However, there is a supergiant, IRS 7, with a magnitude of ∼ 6.5 in the K band, located less than 6 from Sgr A*. IRS 7 is an ideal reference source for NAOS’ infrared wavefront sensor. Combined with the location of the VLT in the Chilean Atacama, where the GC passes close to zenith and allows long observations at low airmasses, NACO is the ideal new generation instrument for observations of the nuclear star cluster of the Milky Way.
The work presented here is based largely on the observational data obtained with the new NACO instrument, which have brought signiﬁcant progress to our knowledge of the nuclear stellar cluster and the nature of Sgr A*. They allowed an examination of the overall structure of the stellar cluster in the central parsec. In combination with a thorough re-analysis of older speckle imaging data, we present observations and analyses on the stellar dynamics near Sgr A*, which provide new and compelling evidence that this source is indeed a supermassive black hole and allow, for the ﬁrst time, a direct, geometrical determination of the distance to the GC. We conclude with the ﬁrst observations of a near-infrared counterpart of Sgr A*, strong evidence for accretion and emission processes near the black hole.
The Central Parsec of the MilkyWay