Jobs / Thesis
The center of the Milky Way
Galactic Center (GC)
is the closest galactic nucleus, and harbors the currently
best (supermassive) black hole (BH) candidate. Present measurements of the proper
motion of the radio source Sgr A* and stars in close orbit around Sgr A*,
in conjunction with basic physical theory, make a compelling case that Sgr A*
indeed must be a 3.5 million solar mass black hole, beyond any reasonable
doubt. However, the final observational proof that Sgr A* has an event
horizon is still lacking. Given its mass, its equivalent zero-spin event
horizon radius (Schwarzschild radius RS) is 9 microarcsec.
Because of its proximity, this is also by far the largest angular size
of any astrophysical black hole, including all known external galaxy central
BH candidates. The Galactic Center is therefore the best laboratory in
astronomy for exploring, by spatially resolved dynamics,
the vicinity of black hole systems. The ultimate goal for GRAVITY is to
probe the space-time around Sgr A* down to a few RS.
We aim at four main scientific topics:
- Determining the nature of the flares of the supermassive black hole
Sgr A* exhibits outbursts of infrared, X-ray, and submillimeter emission typically a
few times a day. These flares last for about 1 h, and their light-curves show significant
variations on a typical timescale of 15 - 20 min in the IR. Given the typical rise
time of the substructures in the light-curve, they must come from regions smaller than about
10 light-minutes, or 17 RS. Recent evidence from multi-wavelength studies and infrared
polarimetry strongly suggests that the emitting region in the near infrared (NIR)
consists of a compact spot of hot gas (emitting synchrotron radiation),
probably in combination with a somewhat more extended and perhaps expanding region.
Such a spot cannot remain static in the potential well of the BH, but its velocity
must be comparable to the Keplerian circular velocity.
Therefore, measuring the 2D astrometry of flares with 10 microarcsec accuracy and a
time resolution of a few minutes will not only allow the determination of the location
of the flares with respect to the BH, but also their proper motions. This will enable
understanding their nature and constraining the processes at work in other galactic nuclei
- Probing the space-time around Sgr A*
Additionally, the flares are a dynamical probe in the potential well of Sgr A*
at a few to 100 RS. If the observed 17 min quasi-periodicity seen in integrated
light in several of the infrared flares and in polarized light in one recent flare is
caused by a hot spot near the last stable orbit of a Kerr BH, it is possible to
deduce the spin parameter. If a simple orbiting hot spot model is indeed
applicable (to some of the flares), then 10 microas-accuracy astrometry may be a
clean enough dynamical probe of the space-time down to the photon-sphere of the
BH In this instance, GRAVITY may then test General Relativity in its strong field limit.
- Probing stellar dynamics in the regime of General Relativity
The current best estimates of the mass of the central BH and the distance to the
GC are obtained through orbit fitting of stars in the central arcsecond of the Galaxy,
the so-called S-stars. Stellar counts predict that a few faint stars
(17.5 < mK < 19.5) should reside even closer to the BH, within the central 100 mas of
the Galaxy. These cusp stars have orbital periods of order one year,
periapses of order 1 mas = 100 RS, and travel at relativistic velocities during
their periapse passages. The repeated interferometric imaging of these
stars will allow testing of relativistic effects, in particular the prograde
periastron shift. In addition, it will give the mass enclosed within 100 RS,
while the currently known S-stars measure the mass enclosed within 1000 RS:
comparing the two numbers will give a measurement of the mass of the stellar cusp.
- Astrometry of S-stars
The astrometric capability of GRAVITY will also allow deriving very accurate
orbits for the more distant S-stars, which are currently followed by the adaptive optics
imaging program at the VLT. Specifically when they pass periapse,
it is possible to get a significant improvement on the determination of the position,
mass, and distance of Sgr A*. These improvements are indeed required to further
constrain the modeling in the two aforementioned experiments. On a longer (~ 10 year)
time-scale, the astrometry of the S-stars will also allow determining up to
second order effects in v/c due to Special and General Relativity, and will
simultaneously probe the extended mass component at the scale of one arcsecond.
Intermediate mass black holes
There is general consensus that MBHs have now been detected in galactic nuclei
down to a mass level of a few 100000 Msun. It is less clear whether intermediate
mass black holes (IMBHs: a few 100 to a few 10000 Msun) also form, for instance by
core collapse in dense stellar clusters. Such IMBHs may also be of relevance as
seeds of the MBH population, perhaps formed at z > 10 in dense population III star clusters.
In the GC, there is now compelling evidence for a very compact cluster (GCIRS 13E) a
few arcseconds away from Sgr A*, which may contain a > 1000 Msun IMBH.
Interferometric imaging of the cluster will provide reliable proper motions
for the IRS13 core stars, and hence dynamically test the IMBH hypothesis.
Globular clusters are other obvious places to look for IMBHs, as are dense,
massive star clusters, such as the Arches cluster near the GC. The current main
limitation on the ability to measure the mass of these putative dark
objects is the small number of bright stars in the core of Globular
Clusters suitable for radial velocity measurements. GRAVITY will dramatically improve
the situation by enabling high accuracy proper motion measurements. At 8 kpc,
1 km/s corresponds to 26 microarcsec/yr, such that this accuracy will be reached
in 1 yr with narrow-angle astrometry, or a few years through imaging.
In addition, the acceleration of a star orbiting a 1000 Msun BH at 4 mpc
(100 mas at 8 kpc) is roughly 7 microarcsec/yr2, so that, for such stars,
accelerations (and orbits) could be detected after a few years in astrometric mode.
Stellar orbits around extragalactic supermassive black holes
The nucleus of M31 is made of a blue disk of 200 Myr old stars orbiting
a 1.4 x 10^8 Msun BH. The half-power radius of this disk is 60 mas = 0.2 pc,
and it should contain of order 10 bright red evolved stars, which would be
resolvable by GRAVITY. If M31 were observable from Paranal, their proper motion,
of order 1700 km/s = 0.5 mas/yr at 0.76 Mpc, would be measurable within only
a few years by VLTI imaging. The same proper motion measurements will be
possible in galaxies with a > 10^7 Msun BH out to a distance of 10 Mpc,
and with a black hole of a few 10^8 Msun or more out to about 30 Mpc, as long as a
few very bright stars can be interferometrically isolated in the central regions.
These constraints potentially make stellar proper motion studies possible
for many tens of known black hole candidate systems. Compared to other
instruments with strong spectroscopic capabilities, GRAVITY will be
superior for this project thanks to its optimization for single-band low
Active Galactic Nuclei (AGN)
The formation and the evolution of SMBHs in galaxies is one of the major
open questions of extragalactic astrophysics. In particular, BH growth and AGN feedback
on the surrounding galaxy host are two key processes that can profoundly affect
galaxy evolution models. In that framework, GRAVITY aims at constraining the size and
composition of the core AGN structures (broad line region (BLR), dusty torus,
condensation region; see figure). Thanks to its high sensitivity (optimized throughput)
and its infrared wavefront sensors, GRAVITY will extend the sample of sources,
compared to other VLTI instruments, therefore providing a greater variety of geometries,
a key in the discussion of the unified model. The four-telescope astrometric
capability will allow the dynamical study of the compact core, especially at an
accuracy of the size of the BLR and even of the outer regions of the accretion disk
for the closest objects.
Many active galactic nuclei are deeply dust-embedded, and therefore faint in the
optical. For these objects the IR wavefront-sensing capability of GRAVITY is mandatory.
High spectral resolution is not necessary for many projects, because the spectral
features of AGNs are comparably broad. A key requirement is the high sensitivity
provided by GRAVITY thanks to the optimization for single band operation at
low spectral resolution. Sources with mK = 12 will be within range for GRAVITY in a
fringe coherencing mode. This is one magnitude better than
(VLT INterferometer Commissioning Instrument). For this reason, GRAVITY will be able
to image several ten AGNs without a nearby phase reference star, which would be
inaccessible to the other less sensitive instruments.
GRAVITY will therefore play a major role in the study of AGNs.
It will test key aspects of the standard unified model by imaging the dust
and molecular torus. Until recently, the size of the putative torus was
measured from the distance at which graphite and silicate grains can
condense, or from the modeling of the IR spectral energy distribution.
The size of the torus of NGC 1068 has been recently directly measured with
MIDI (MID-infrared Interferometric instrument at VLTI.
GRAVITY will also be able to study the broad line region, a compact region in
which gas streaming velocities reach several thousand km/s. Up to now,
the BLR size has been estimated from reverberation mapping measurements
in the blue part of the visible spectrum. Near infrared observations are
sensitive to both the dust torus and the more compact central source.
Spectroscopically resolved observations will allow disentangling the emission from
regions with spectral features such as the BLR, and continuum sources such as the
dusty torus or the central engine. The largest BLRs are within reach at
NIR wavelengths with angular scales of 4 mas for sources closer than 100 Mpc.
The application of 10 microarcsec astrometry in velocity resolved mode of GRAVITY
(possible for a spectral resolving power of 500 in the BLR) will allow,
for the first time, the direct measurement of the spatial extent of the
BLR and an investigation of the degree of ordering of the gas motions
(e.g. rotation). In addition, the combination of imaging and spectroscopy,
will yield, through tomography,
the 3D structure of the BLR.
Masses of the most massive stars
There still exists a discrepancy of up to a factor of two in the mass
estimates for the most massive main-sequence stars. Comparing spectra with atmospheric
models yields upper mass limits in the range of 60 Msun, whereas evolutionary
tracks and observed luminosity suggest a mass of up to 120 Msun for stars of spectral
type O2V and O3V. Clearly, dynamical mass estimates for early O-type main-sequence
stars are required. Luckily, quite a number of spectroscopic binary O-stars
are known in the cores of Galactic starburst clusters like the Arches, the Quintuplet and
NGC 3603, or extragalactic starbursts like 30 Doradus. With a nominal
resolution of 4 mas at a wavelength of 2 micron, GRAVITY could resolve some of
the longer period spectroscopic binaries, and monitor the astrometric motion
of the photo-center for the shorter period, closer binaries with a precision
of 10 microarcsec. Astrometric orbits for these deeply embedded binary stars
will hence directly yield dynamical mass estimates. The unique narrow angle
astrometry mode of GRAVITY is ideal for these dynamical studies in crowded
regions. Several objects (e.g. the Arches, the Quintuplet) cannot be observed
without the infrared wavefront sensing provided by GRAVITY (Figure 4).
Circumstellar disks and jets around young stars
Circumstellar disks and outflows are closely linked to the star formation process.
The presence of a circumstellar disk is also the prerequisite for the formation of
planetary systems. The gravitational interaction between a planet and a disk should
manifest itself by the occurrence of spiral structures, wakes and gaps. Such structures
have indeed already been observed in a number of cases (e.g. GG Tau, Formalhaut, etc.).
The relative faintness of a young giant planet compared to the high surface brightness
of a typical circumstellar disk thus far prevented its direct detection in diffraction
limited observations with 8m class telescopes in the near infrared. GRAVITY's 4 mas
resolution (compared to 60 mas resolution for one UT) drastically improves the
contrast between a disk and its embedded planet by a factor of (60/4)^2 = 225. Hence
it should be possible to probe for young giant planets, which are almost 6 mag fainter
than what is currently achievable. Indirect evidence for the presence of such planets
could be derived from the photocenter movement of the exoplanet host star around the
common center of mass. The processes occur around young, embedded stellar systems;
therefore GRAVITY's infrared wavefront-sensing capability is critical.
While the ubiquity of jets in star forming regions has been well established,
the physics behind the formation of jets, and in particular the launching mechanism,
is still poorly understood. Models (Figure 5) suggest that angular momentum gets first
transferred along horizontal magnetic field lines from the disk to the central material,
which then gets accelerated along a vertical pressure gradient, ultimately forming a
collimated jet. The important processes seem to take place within less than 0.5 AU
from the star, which at typical distances to the nearest star forming regions of 150 pc
translates into an angular size of less than 30 mas. At 4 mas resolution, GRAVITY will
be able to resolve the central jet formation engine around young, nearby stars.
Furthermore, at a distance of 150 pc, an astrometric precision of 10 microarcsec over
a time span of 1 h corresponds to a transversal velocity accuracy of 60 km/s. Hence,
high-velocity outflows and the formation and evolution of jets from TTauri stars
with typical velocities of 150 km/s can be resolved and traced in real time by
GRAVITY. These observations will put tight constraints on jet formation models
and the role of magnetic fields.
© Infrared and Submillimeter Astronomy Group at MPE
14/03/2006, editor of this page: Stefan Gillessen