Galactic Center
Max-Planck-Institut für extraterrestrische Physik

- Infrared/Submillimeter Astronomy -

Galactic Center Research

 .MPE . Infrared Astronomy . Research . Galactic Center

Galactic Center

Image Gallery



Valid HTML 4.01!



Moving stars

Since 1992 we have observed the central parsec of the galactic center in K-Band. We mainly use our Speckle-Camera SHARP I on the NTT located at La Silla, Chile, and NAOS/CONICA at the VLT on Paranal, which are operated by the European Southern Observatory.

Our primary goals are to find out
  • whether there is a massive black hole
  • when and where the stars were born
  • the nature of these stars
  • the dynamics in this region

So far, we have shown that at least two periods of star formation have occured. There are three sorts of stars in the galactic center:

  • Stars with He I - emission
  • Stars which show CO - absorption
  • Embedded stars with featureless spectra

WARNING - The movie to the right is quite big (roughly 6MB). If you wish, you can stop the download and download a lightweight MPEG-version.

Radio and X-ray emission from Sgr A* Infrared flare

The supermassive black hole in the centre of the Milky Way was discovered as a bright non-thermal radio source in the 1970s and termed Sagittarius A* (Sgr A*). Potential X-ray radiation by Sgr A* was detected with the X-ray observatory Rosat in the 1990s. A reliable identification of X-rays from Sgr A* was finally possible with the new X-ray satellites Chandra and XMM, with their high spatial resolution and sensitivity. The radio emission of SgrA* only varies slowly on time scales of several days to a few hundred days and generally with an amplitude <10%. However, in the X-ray regime, SgrA* was found to exhibit two different states. On the one hand, in the quiescent state, weak X-ray emission appears to come from a slightly extended area around the black hole that appears to be evidence of hot accreting gas in the environment of SgrA*. On the other hand, SgrA* shows X-ray flares with a period of about one per day. During these flares, the emission rises by factors up to 100 during several tens of minutes and a distinctive point source becomes visible at the location of SgrA*. The short rise-and-decay times of the flares suggest that the radiation must origin from a region within less than 10 Schwarzschild radii of a 3.6 million solar mass black hole.

Near-infrared flares from the black hole

Near-infrared high-resolution observations of the galactic centre (GC) became possible since the beginning of the 1990s. Since then, the GC stellar was regularly monitored by high-resolution NIR imaging. However, in spite of all efforts, no unambiguous NIR counterpart of SgrA* could be detected up to 2003. On the 9th of May, during routine observations of the GC star cluster at 1.7 microns with the CONICA/NAOS adaptive optics imager/near infrared camera at the ESO VLT, we witnessed a powerful flare at the location of the black hole. Within a few minutes, the flux of a faint source increased by a factor of 5-6 and fainted again after about 30 min. The flare was found to have happened within a few milli-arcseconds of the position of Sgr A*. The short rise-and-decay times told us that the source of the flare was located within less than 10 Schwarzschild radii of the black hole.

Detection of IR radiation from Sgr A*

Fig 1: Detection of NIR emission from Sgr A*. The images show raw AO images (60 s total exposure time) of an area 1''x1'' around Sgr A*, observed on May 9, 2003 UT. The left image was taken at the beginning of the observations, the right image about 40 min later. The flaring source is easily detected in the right image. Its position is offset -1.4±3.0 mas in R.A. and -0.2±3.0 in Dec. from the dynamical position of Sgr A* as it was determined from the orbit of the star S2. The star S2 is marked by a cross, the position of Sgr A* is indicated by a white circle.

During subsequent observations in 2003, we could observe more flares from Sgr A* and also quiescent emission from a source at this location. With hindsight, we could also detect a flaring source in older, longer wavelength data from 2002. At the moment, we have observed a total of four flares in the H, K and L-bands (1.7, 2.2 and 3.8 microns). The flares were observed at four different epochs within a few milli-arcseconds of the location of SgrA*, which makes it highly probable that they are indeed associated with matter in the immediate environment of the black hole, which is also reflected in the very short rise-and-fall time scales of the light curves. Independently, flaring and variability of SgrA* in the L-band was also observed at the Keck telescope by researchers from the University of California, LA, in June 2003.

Light curves of Sgr A* NIR flares

Fig 2: Light curves of the Sgr A* NIR flares in 2002 and 2003, observed with NACO/VLT. The L'-band flare on August 30, 2002, was only partially covered by observations. Gaps in the time series of the H-band flare on May 10, 2003, and of the KS-band flare on June 15, 2003, are due to sky observations and instrument failure, respectively. For comparison, the emission of the steady emission of the star S1 near Sgr A* is shown in all the plots (light grey data points). Arrows in the plots of the two KS-band flares indicate substructure peaks of the flares. Both KS-band flares show very similar quasi-periodicity, although the second flare was observed more than 24 h after the first one and must thus have been an unrelated event. The upper right panel shows the normalised power spectrum of the two KS-band flares. Both of them show a significant peak at a frequency corresponding to time scales of 16.8±2.0 min. In both cases, the power spectrum of S1 does not show such a frequency.

The quiescent and flaring NIR emission from SgrA* fills an important gap in our knowledge of the spectrum of this source and will allow to constrain the existing models on how the radiation is produced. While the quiescent emission appears to be largely consistent with an origin in the high-energy tail of a synchrotron spectrum, the mechanism of the NIR flares is uncertain. Although the NIR flares were observed at different epochs, they might hint at a blue colour of the flares which would be a challenge to current theories. Simultaneous, Multi-wavelength NIR and X-ray observations of the GC are planned for the next year. The chances are high that these observations will provide the required data to constrain the models and to establish (or exclude) a relation between the X-ray and NIR variability.

Spectral Energy Distribution of Sgr A*

Fig 3: Spectral energy distribution of the emission from Sgr A*. This plot shows the extinction and absorption corrected luminosities. All error bars are ±1 sigma and include statistical and systematic errors. Black triangles denote the radio spectrum of Sgr A*. Open grey circles mark various infrared upper limits from the literature. The three X-ray data ranges are (from bottom to top) the quiescent state as determined with Chandra (black; Baganoff et al., 2003), the autumn 2000 Chandra flare (red; Baganoff et al., 2000), and the autumn 2002 flare observed by XMM (light blue; Porquet et al., 2003). Open red squares with crosses mark the de-reddened peak emission (minus quiescent emission) of the four NIR flares. Open blue circles mark the de-reddened H, KS, and L' luminosities of the quiescent state, derived from the local background subtracted flux density of the point source at the position at Sgr A*, thus eliminating the contribution from extended, diffuse light due to the stellar cusp around Sgr A*.

A spin measurement of the black hole?

The two K-band flares observed on the 15th and 16th of June 2003 are the flares that were completely covered by observations. Although they happened more than 24 hours apart and thus appear to be unrelated events, they both show a striking quasi-periodicity of the flare with a period of about 17 min. Of all possible periodic processes near a black hole (acoustic modes of a thin disk, Lense-Thirring precession, precession of orbital nodes, orbital motion), the period of matter circling the black hole near the last stable orbit is the shortest one. The observed period of 17 min is so short, however, that the only reasonable explanation is that the oscillations are produced by Doppler boosting of hot gas near the last stable orbit of a spinning (Kerr) black hole. The spin of the black hole will allow for a last stable orbit closer to the event horizon and thus with a shorter orbital frequency. From the observed 17 min period we estimate that the supermassive black hole Sgr A* has a spin that is half as big as the maximum possible spin of such an object.

Additional observations of flares and their quasi-periodicity will be needed in order to confirm this result. Should the quasi-periodicity indeed be an intrinsic feature of the flares then this will mean that the era of black hole physics has begun with the properties of black holes accessible to direct measurements!

© Infrared and Submillimeter Astronomy Group at MPE
last update: 07/10/2002, editor of this page: Thomas Ott


up © Max-Planck-Institut für extraterrestrische Physik