Sgr A

19.3.1 X-Ray Detection of Sgr A*

After 20 y searching for high energy emission from Sgr A* [19,27], it has come to a turning point when Chandra Observatory with its 0.5" resolution could detect weak emission from the radio source position [2]. Surprisingly, Sgr A* is much fainter than expected from accretion onto a super-massive black hole. In particular, in the 2-10keV energy band its (quiescent) X-ray luminosity is only about 2.2 x 1033 ergs_1 within a radius of 1.5" [3]. This value may in fact be considered as an upper limit since this region contains other components such as stars, hot gas, etc. Thus, Sgr A* radiates in X-rays at about 11 orders of magnitude less than its corresponding Eddington luminosity.

19.3.2 Flaring Sgr A*

Then, during a Chandra observation in October 2000, the same source was seen to flare up by a factor of «50 within 3 h [2]. The flare had a duration of about 10 ks, with L(2-10keV) =1.0 ± 0.1 x 1035 erg s^1 for the flare peak. The discovery



6000 8000 Time (s)




Fig. 19.6 XMM-Newton X-ray light curve (EPIC MOS 1+MOS2+PN) within a radius of 10" around the Sgr A* position. The time binning is 100 s, and the error bars indicate 1c uncertainties. The 2-10keV light curve shows the quiescent and the flare periods [18]

of X-ray flares from Sgr A* has provided new exciting perspectives for the understanding of the processes at work in the galactic nucleus.

A second significant flare was detected with XMM-Newton [7]. It showed a monotonic flux rise up to a factor of about 20-30 in the last 900 s of the observation. As in the Chandra Observatory observation, the spectrum was rather hard with a power law photon index r = 0.9 and a maximum observed 2-10 keV luminosity of about 6 x 1034 erg s^1. On the basis of and further flare detections, the flare rate could be estimated to be slightly more than on flare per day.

Then, in October 2002, the brightest flare so far was detected with XMM-Newton. With a peak luminosity of almost 4 x 1035 erg s^1, it was 160 times brighter than at quiescent level [18], Fig. 19.6. Remarkably for this flare was its extremely short duration of less than 1 h and its rather soft spectrum with r « 2.5.

19.3.3 The Nature of Sgr A*

A detailed analysis of this and the preceding flares revealed quasi-periodic oscillations with several distinct periods between «100 and 2000 s, which were assumed to be characteristic cyclic modes of the accretion disk. From these modes, both the mass and the angular momentum of the central black hole could be derived [1].

Since also in the near-infrared regime Sgr A* could be detected first by its flaring behavior, it was quite natural to search for simultaneous flares both in NIR and X-rays. Success came finally in 2004 using the NACO adaptive optics instrument at the Very Large Telescope of the European Southern Observatory and the Chandra Observatory [6].

Current models that explain the Sgr A* spectral energy distribution invoke radi-atively inefficient accretion flow models, RIAFs [16] including advection dominated accretion flows, ADAF [15], convection-dominated accretion flows, ADIOS [4] or Bondi-Hoyle accretion [11] and jet-models [10]. The recent simultaneous X-ray/NIR detection of the Sgr A* counterpart suggests that at least for the observed flare it is the same population of electrons that is responsible for both the IR and the X-ray emission, regardless of the emission mechanism. While it is not yet possible to completely rule out any of the proposed models, it is found that an attractive mechanism to explain the observed simultaneous NIR/X-ray flare is the synchrotron self-Compton (SSC) process. In this model, the X-ray photons are produced by up-scattering of millimeter or submillimeter photons [6].

Doppler boosting could quite conveniently explain the flaring behavior of Sgr A*. Doppler boosting will occur in models that involve relativistic outflows or jets pointing toward the observer at a small angle to the line of sight [10]. In the context of this jet model, the emitting component would be located close to the jet base and would have a size of a few Schwarzschild radii or less.

0 0

Post a comment