4. X-rays from the Sgr A Complex
The location of SgrA* was imaged on August 24-26, 1997 for a net exposure time of 99.5 ks. In order to study the surface brightness and temperature profile of the diffuse emission, we extracted the MECS counts from four annular regions with different inner and outer radii (,,,) centered on SgrA*. All these spectra contain several emission lines, with the K-lines from iron (E keV) and sulfur (E keV) particularly bright. We fitted them with a single temperature plasma model (MEKAL in XSPEC v.10.00), deriving the surface brigthness and temperature profiles shown in Fig. 2. A radial spectral variation is evident: while the keV temperature in the three external regions is almost constant, the emission from the inner is significantly softer. This is probably due to the presence, in addition to the diffuse emission, of further contributions from the bright point sources present near the GC (Maeda et al. 1996, Predehl & Trümper 1994, Sidoli et al. 1999b).
The fit with a single temperature thermal model does not account for all the emission lines at low energies (in particular the intense emission from sulfur at E2.4 keV), nor for an excess around 6.4 keV, due to the presence of emission of fluorescent origin. A gaussian line added at 6.4 keV accounts for these residuals. While this line is present in all the spectra extracted from to , it seems to be absent within (EW12 eV). This is consistent with the 6.4 keV map produced with ASCA (Maeda & Koyama 1996) where a strong peak is visible in the region of the Radio Arc, NE to the GC. Indeed the EW of this line increases towards the outer regions ( eV, eV, eV). The MECS spectra extracted from two semi-annular regions at NE and SW of the GC, confirm this interpretation: the EWs of the 6.4 keV line are eV (NE sector; errors are at 90% confidence level) and eV (SW sector).
Since the temperature profile does not show evidence of strong spectral variations in the region from to , we analysed the overall spectrum by extracting all the counts from this larger circular corona. We started by fitting a thermal bremsstrahlung plus three gaussian lines at 1.8, 2.4 and 6.7 keV (Si, S and Fe respectively). Their estimated equivalent widths are about 120, 190 and 1000 eV. The best fit value for the bremsstrahlung temperature is 13 keV. This temperature is too high to be consistent with the presence of the low energy emission lines (Kaneda et al. 1997, Fig. 2b). A possible explanation is a multi-temperature plasma. Thus we fitted the spectrum with two thermal emission plasma models (two "MEKAL" in XSPEC). Our best fit parameters (=1.29, 368 d.o.f) are NH= cm-2, keV and keV. A gaussian added at 6.4 keV gives an EW120 eV. The total unabsorbed flux (2-10 keV) is ergs cm- 2 s-1, about one third of which is contributed by the soft component.
To study the spatial distribution of the diffuse emission, we extracted images in different energy bands (Fig. 3). The images corresponding to the 2-5 keV and 7-10 keV ranges show significantly different distributions of the diffuse emission.
Both the soft and hard emissions are peaked at the GC position, but they have different spatial extents. In particular, the softer emission has a nearly triangular shape, with a rather sharp decrease in the south-western side, which is absent in the hard X-ray map. To study the spatial distribution of the iron line we have also extracted an image in the 5.5-7.5 keV band and subtracted from this map the continuum emission interpolated from the contiguous energies. The resulting iron line image shows a spatial distribution elongated along the galactic plane, very similar to that of the hard X-ray map.
© European Southern Observatory (ESO) 1999
Online publication: September 2, 1999